commit 9cf2676b8cf9a35d7514819dc0985376eff915d9
Author: ModelHub XC <noreply@modelhub.org.cn>
Date:   Wed Aug 26 06:52:13 2026 +0800

    初始化项目，由ModelHub XC社区提供模型
    
    Model: luoooo/llama_ft_42_2523
    Source: Original Platform

diff --git a/.gitattributes b/.gitattributes
new file mode 100644
index 0000000..7bc225d
--- /dev/null
+++ b/.gitattributes
@@ -0,0 +1,34 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bin.* filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zstandard filter=lfs diff=lfs merge=lfs -text
+*.tfevents* filter=lfs diff=lfs merge=lfs -text
+*.db* filter=lfs diff=lfs merge=lfs -text
+*.ark* filter=lfs diff=lfs merge=lfs -text
+**/*ckpt*data* filter=lfs diff=lfs merge=lfs -text
+**/*ckpt*.meta filter=lfs diff=lfs merge=lfs -text
+**/*ckpt*.index filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
\ No newline at end of file
diff --git a/config.json b/config.json
new file mode 100644
index 0000000..d32102a
--- /dev/null
+++ b/config.json
@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "huggyllama/llama-7b",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 2048,
+  "max_sequence_length": 2048,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 32,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.31.0",
+  "use_cache": false,
+  "vocab_size": 32000
+}
diff --git a/configuration.json b/configuration.json
new file mode 100644
index 0000000..f9291c3
--- /dev/null
+++ b/configuration.json
@@ -0,0 +1 @@
+{"framework":"Pytorch","task":"text-generation"}
\ No newline at end of file
diff --git a/generation_config.json b/generation_config.json
new file mode 100644
index 0000000..b63f378
--- /dev/null
+++ b/generation_config.json
@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.31.0",
+  "use_cache": false
+}
diff --git a/global_step2523/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..ffb98ef
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f58bb532bc515f3168a8d9c899900ff846f7997e428bf8cbaf1a7ea570d5c3b
+size 10107628951
diff --git a/global_step2523/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..ec2335e
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb5d8f4338f09924db19be90963d343126bb0d09d5de3432303d16f73a0335bf
+size 10107629335
diff --git a/global_step2523/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..c2f77f8
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9fe1eea24d86c44e768ccc321776bc178efe5875e8b392baf48d342c1b8e900
+size 10107629335
diff --git a/global_step2523/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..3edfe6b
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be8ab0f68e1b58942c184bed04b664a604d051f9c47f6c3077ffb05ff15e36f3
+size 10107629527
diff --git a/global_step2523/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..15c16e3
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03315647fdcf8c6aa6bb363a680e9b54143e7baeb8c362e743a54bdd614bc40c
+size 10107629399
diff --git a/global_step2523/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..85f8462
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1d069e7d978ad33bf47f6bfa68de514adb52138cd06c8df45b5092370428b9b
+size 10107629463
diff --git a/global_step2523/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..f370ec2
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0178fc300026585f803996e27d5aa3c65b570df50dc2edeb808c4c4d66642b69
+size 10107629335
diff --git a/global_step2523/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt b/global_step2523/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt
new file mode 100644
index 0000000..beda357
--- /dev/null
+++ b/global_step2523/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:389e76abe60ba555040de17acb473d7a45292e7cd9f3061efa1e52e9984097c0
+size 10107629015
diff --git a/global_step2523/mp_rank_00_model_states.pt b/global_step2523/mp_rank_00_model_states.pt
new file mode 100644
index 0000000..8ecf254
--- /dev/null
+++ b/global_step2523/mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c71d219f3ce6a8054ebfb97b76b1e9559db2bb29b6fecce638d5c473ab518664
+size 13476935962
diff --git a/latest b/latest
new file mode 100644
index 0000000..aee5fa3
--- /dev/null
+++ b/latest
@@ -0,0 +1 @@
+global_step2523
\ No newline at end of file
diff --git a/pytorch_model-00001-of-00002.bin b/pytorch_model-00001-of-00002.bin
new file mode 100644
index 0000000..be6f7a2
--- /dev/null
+++ b/pytorch_model-00001-of-00002.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ba840d6b0a6a34218e46b3f4f71b0d55297df0a53c2f816a74e33c2f3d7ad86
+size 9976631486
diff --git a/pytorch_model-00002-of-00002.bin b/pytorch_model-00002-of-00002.bin
new file mode 100644
index 0000000..32a2205
--- /dev/null
+++ b/pytorch_model-00002-of-00002.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f0bd8e1dfb7e55621d8587035f38b1eb2f088827c334fc7e7c7e2818404926c
+size 3500314451
diff --git a/pytorch_model.bin.index.json b/pytorch_model.bin.index.json
new file mode 100644
index 0000000..77997ab
--- /dev/null
+++ b/pytorch_model.bin.index.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f56e69fedde5d28e4f37f2b62f74e8522bbfa13395a6d696d1ef99222a431ab7
+size 26788
diff --git a/rng_state_0.pth b/rng_state_0.pth
new file mode 100644
index 0000000..6ab8e21
--- /dev/null
+++ b/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a118cdad874e46838801d9c7548cab04dc2a590c59d5cdd5d061e814500301ed
+size 21687
diff --git a/rng_state_1.pth b/rng_state_1.pth
new file mode 100644
index 0000000..c1cd007
--- /dev/null
+++ b/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae9d159f9d8e01065d2faf211aa451a41fa03cf513d15d96ed19781625f0f17d
+size 21687
diff --git a/rng_state_2.pth b/rng_state_2.pth
new file mode 100644
index 0000000..01c23f1
--- /dev/null
+++ b/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdaf5a1f643d3d172c1f22702aa6ef7a128eb925a89ff542ad832a7b5d119173
+size 21687
diff --git a/rng_state_3.pth b/rng_state_3.pth
new file mode 100644
index 0000000..e286ec3
--- /dev/null
+++ b/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bd1bb67ec48f262adbb9ab9ca45e647cef2a853f78095fd748c47258c0fefe3
+size 21687
diff --git a/rng_state_4.pth b/rng_state_4.pth
new file mode 100644
index 0000000..abf296f
--- /dev/null
+++ b/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fff3be49764eac7f30ddf56707cca6516b91fe343c27fd8739f3ee2f6c096833
+size 21687
diff --git a/rng_state_5.pth b/rng_state_5.pth
new file mode 100644
index 0000000..4da97e6
--- /dev/null
+++ b/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8659227dd6e0b64c6f71c64b0f35092a523f4f7660d545ca53b8281a1d230167
+size 21687
diff --git a/rng_state_6.pth b/rng_state_6.pth
new file mode 100644
index 0000000..911d6f9
--- /dev/null
+++ b/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f36b5a916ec825a4e78aed91ab2b13b2c8e006ccc75cabb6e74078d41f89369
+size 21687
diff --git a/rng_state_7.pth b/rng_state_7.pth
new file mode 100644
index 0000000..816883f
--- /dev/null
+++ b/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0d61140cb3373aa861cbf10ea5532af699368a670d67b8ce275c2ac09e848f7
+size 21687
diff --git a/special_tokens_map.json b/special_tokens_map.json
new file mode 100644
index 0000000..f928b24
--- /dev/null
+++ b/special_tokens_map.json
@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}
diff --git a/tokenizer.model b/tokenizer.model
new file mode 100644
index 0000000..6c00c74
--- /dev/null
+++ b/tokenizer.model
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723
diff --git a/tokenizer_config.json b/tokenizer_config.json
new file mode 100644
index 0000000..60bcd0d
--- /dev/null
+++ b/tokenizer_config.json
@@ -0,0 +1,35 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "legacy": true,
+  "model_max_length": 1280,
+  "pad_token": null,
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}
diff --git a/trainer_state.json b/trainer_state.json
new file mode 100644
index 0000000..ba3950f
--- /dev/null
+++ b/trainer_state.json
@@ -0,0 +1,15154 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "global_step": 2523,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 2.0503,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 2.3069,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.9309,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.7445,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.8824,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.9035,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 2.2855,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.9347,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.7086,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.8894,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.8565,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.9346,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.8632,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.8608,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.8239,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.8494,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 2.0824,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.3544,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.9392,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.7509,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.2608,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.3936,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.4369,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.8616,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.7469,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.3212,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.2939,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.2779,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.2412,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.7532,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.2469,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.3184,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.7245,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.8052,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.779,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.3537,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.2419,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 2.0561,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.156,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.7532,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 2.0618,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.8058,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.3908,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 2.0388,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.8684,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.8312,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.9905,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 2.0442,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.1189,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.9317,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.8889,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.9161,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.6701,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.7986,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.3986,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.9346,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.7632,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.8013,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.9406,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 2.0405,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.7689,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.8909,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 2.1128,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.3193,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.8801,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 113
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 114
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.7817,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.7488,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.2364,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.7542,
+      "step": 119
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.0904,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.9077,
+      "step": 121
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 122
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 123
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 124
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.8111,
+      "step": 125
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.6876,
+      "step": 126
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 127
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 128
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.8221,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.692,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.2236,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.2822,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 138
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.9747,
+      "step": 139
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.8034,
+      "step": 141
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.9027,
+      "step": 142
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 143
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 144
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.1157,
+      "step": 145
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.3757,
+      "step": 146
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 147
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 148
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 149
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.805,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 151
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.4192,
+      "step": 152
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 153
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 154
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.8891,
+      "step": 155
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 2.1525,
+      "step": 156
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 157
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.848,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.817,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.3342,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.784,
+      "step": 162
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 163
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 164
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.4392,
+      "step": 165
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 166
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 167
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.4113,
+      "step": 168
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.9013,
+      "step": 169
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.3502,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.3315,
+      "step": 171
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.8155,
+      "step": 172
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.7438,
+      "step": 173
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 174
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 2.149,
+      "step": 175
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.2719,
+      "step": 176
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 177
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 2.0176,
+      "step": 178
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.369,
+      "step": 179
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.9747,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 181
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 182
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 183
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.9984,
+      "step": 184
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 185
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 186
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.8617,
+      "step": 187
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.9649,
+      "step": 188
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 1.3607,
+      "step": 189
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.8665,
+      "step": 190
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 2.0313,
+      "step": 191
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 192
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 193
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.7192,
+      "step": 194
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 195
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 196
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.8236,
+      "step": 197
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.38,
+      "step": 198
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.1611,
+      "step": 199
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 2.0245,
+      "step": 201
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.9086,
+      "step": 202
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 203
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 204
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 205
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 206
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.9472,
+      "step": 207
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 208
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.333,
+      "step": 209
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.811,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.5058,
+      "step": 211
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 212
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.4431,
+      "step": 213
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001,
+      "loss": 1.2619,
+      "step": 214
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 215
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.2802,
+      "step": 216
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.243,
+      "step": 217
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.8506,
+      "step": 218
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.3257,
+      "step": 219
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 221
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.7824,
+      "step": 222
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.3294,
+      "step": 223
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.9603,
+      "step": 224
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 225
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.3964,
+      "step": 226
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.2323,
+      "step": 227
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 228
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 229
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.1652,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.8926,
+      "step": 231
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.7487,
+      "step": 232
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.8996,
+      "step": 233
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 234
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.9859,
+      "step": 235
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 236
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 237
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 238
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 239
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.3188,
+      "step": 240
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.2912,
+      "step": 241
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.2585,
+      "step": 242
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.4188,
+      "step": 243
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.3016,
+      "step": 244
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 245
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.7907,
+      "step": 246
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.9753,
+      "step": 247
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.9289,
+      "step": 248
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 249
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 251
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 252
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 253
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 254
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 255
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 256
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.2452,
+      "step": 257
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.8333,
+      "step": 258
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 259
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.8273,
+      "step": 261
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.8607,
+      "step": 262
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.8191,
+      "step": 263
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001,
+      "loss": 1.294,
+      "step": 264
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 265
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 266
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.2153,
+      "step": 267
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.1439,
+      "step": 268
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 269
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.3432,
+      "step": 270
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.8263,
+      "step": 271
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 272
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.0921,
+      "step": 273
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.2748,
+      "step": 274
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.2216,
+      "step": 275
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.3161,
+      "step": 276
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 2.0079,
+      "step": 277
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 278
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 279
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.3434,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 281
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.3801,
+      "step": 282
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 283
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.8699,
+      "step": 284
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 285
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 286
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 287
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.7456,
+      "step": 288
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.3174,
+      "step": 289
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 290
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.309,
+      "step": 291
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 2.0285,
+      "step": 292
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.306,
+      "step": 293
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 294
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.812,
+      "step": 295
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 2.0005,
+      "step": 296
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 297
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.3901,
+      "step": 298
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 299
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 301
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 302
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 303
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.7878,
+      "step": 304
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.3311,
+      "step": 305
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.2791,
+      "step": 306
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.8455,
+      "step": 307
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 308
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.7448,
+      "step": 309
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.7407,
+      "step": 311
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.7096,
+      "step": 312
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 313
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 314
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001,
+      "loss": 1.466,
+      "step": 315
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.7202,
+      "step": 316
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.2301,
+      "step": 317
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 2.0081,
+      "step": 318
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 319
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 320
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.8195,
+      "step": 321
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.379,
+      "step": 322
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.6832,
+      "step": 323
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.2213,
+      "step": 324
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 325
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.8412,
+      "step": 326
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 327
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.6758,
+      "step": 328
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 329
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.4196,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 331
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.0321,
+      "step": 332
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.8193,
+      "step": 333
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 334
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.7691,
+      "step": 335
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 336
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 2.1256,
+      "step": 337
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 338
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.8938,
+      "step": 339
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001,
+      "loss": 1.3362,
+      "step": 340
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 341
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 342
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 2.0032,
+      "step": 343
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 344
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.3328,
+      "step": 345
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 346
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 347
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 348
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 349
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 2.0005,
+      "step": 350
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 351
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 352
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 353
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.9281,
+      "step": 354
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 355
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.4317,
+      "step": 356
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.9677,
+      "step": 357
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 2.2757,
+      "step": 358
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.8897,
+      "step": 359
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.2708,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.9894,
+      "step": 361
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.4618,
+      "step": 362
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 363
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 364
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 365
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.8071,
+      "step": 366
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.3291,
+      "step": 367
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.4343,
+      "step": 368
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 369
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 370
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 371
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.1842,
+      "step": 372
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 373
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.7701,
+      "step": 374
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 375
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.7753,
+      "step": 376
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 377
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 378
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 379
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.9807,
+      "step": 381
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.7666,
+      "step": 382
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 383
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.4833,
+      "step": 384
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.3147,
+      "step": 385
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 386
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 387
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 388
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 389
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 391
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 392
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 393
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.3081,
+      "step": 394
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.807,
+      "step": 395
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 396
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 397
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 398
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 2.0741,
+      "step": 399
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.7204,
+      "step": 401
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.678,
+      "step": 402
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.4318,
+      "step": 403
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 404
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.9627,
+      "step": 405
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.1563,
+      "step": 406
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 407
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 408
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 409
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.9393,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 411
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.4928,
+      "step": 412
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 413
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 414
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.4432,
+      "step": 415
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 416
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 417
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 418
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 419
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 420
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.1042,
+      "step": 421
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.4346,
+      "step": 422
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 423
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.9813,
+      "step": 424
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 425
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.785,
+      "step": 426
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 2.1507,
+      "step": 427
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 428
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.129,
+      "step": 429
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 430
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.1887,
+      "step": 431
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 432
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.8514,
+      "step": 433
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.7498,
+      "step": 434
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.9916,
+      "step": 435
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 436
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.406,
+      "step": 437
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.8368,
+      "step": 438
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 439
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.3526,
+      "step": 440
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 441
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.257,
+      "step": 442
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.3159,
+      "step": 443
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 444
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 445
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.8515,
+      "step": 446
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.2815,
+      "step": 447
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.4194,
+      "step": 448
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 449
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 450
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 451
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 452
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.8799,
+      "step": 453
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 454
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.803,
+      "step": 455
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.3997,
+      "step": 456
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.956,
+      "step": 457
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 458
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 459
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 461
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 462
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 463
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.2251,
+      "step": 464
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.9854,
+      "step": 465
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001,
+      "loss": 1.3479,
+      "step": 466
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 467
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 468
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 469
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.5459,
+      "step": 470
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 471
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.4589,
+      "step": 472
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.8633,
+      "step": 473
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 474
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.2398,
+      "step": 475
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 2.2392,
+      "step": 476
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 477
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 478
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 479
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.9024,
+      "step": 480
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 481
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 482
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.8927,
+      "step": 483
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 484
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.4375,
+      "step": 485
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 486
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.5093,
+      "step": 487
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.8785,
+      "step": 488
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 489
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 2.5841,
+      "step": 490
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001,
+      "loss": 2.0481,
+      "step": 491
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 492
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 493
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 494
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 495
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.3311,
+      "step": 496
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 497
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 498
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 499
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.9579,
+      "step": 500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.8056,
+      "step": 501
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.4203,
+      "step": 502
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 503
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 504
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.9039,
+      "step": 505
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 506
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 507
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 508
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 2.0212,
+      "step": 509
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 510
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.8438,
+      "step": 511
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.755,
+      "step": 512
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.8514,
+      "step": 513
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.8942,
+      "step": 514
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.9471,
+      "step": 515
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 516
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 517
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.1122,
+      "step": 518
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 519
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 520
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 521
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 522
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 523
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.9642,
+      "step": 524
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 525
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 526
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.3194,
+      "step": 527
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 528
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 529
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 530
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 531
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 532
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 533
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 534
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 535
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 536
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.8462,
+      "step": 537
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.3689,
+      "step": 538
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.3177,
+      "step": 539
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 540
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 541
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 542
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 543
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.3566,
+      "step": 544
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.1639,
+      "step": 545
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 546
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 547
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.7565,
+      "step": 548
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 549
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 550
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 551
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 552
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 553
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.2046,
+      "step": 554
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 555
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 556
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.0604,
+      "step": 557
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 558
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 559
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.6062,
+      "step": 560
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 561
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 562
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 563
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 564
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 2.1036,
+      "step": 565
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 566
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001,
+      "loss": 1.248,
+      "step": 567
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.901,
+      "step": 568
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.3697,
+      "step": 569
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.9608,
+      "step": 570
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.8117,
+      "step": 571
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 572
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 573
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.8145,
+      "step": 574
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.6429,
+      "step": 575
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 2.2943,
+      "step": 576
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 2.1256,
+      "step": 577
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 578
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.9672,
+      "step": 579
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.3168,
+      "step": 580
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 581
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 582
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.7387,
+      "step": 583
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 584
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 2.4101,
+      "step": 585
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 586
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 587
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.7615,
+      "step": 588
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 589
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 590
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 591
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001,
+      "loss": 2.0262,
+      "step": 592
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.8673,
+      "step": 593
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 2.2569,
+      "step": 594
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.9001,
+      "step": 595
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 596
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 597
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 598
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 2.2267,
+      "step": 599
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 601
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 602
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 603
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.9441,
+      "step": 604
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 605
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 2.0828,
+      "step": 606
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 0.9647,
+      "step": 607
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.618,
+      "step": 608
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 2.0988,
+      "step": 609
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.4612,
+      "step": 610
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 611
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 612
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.7284,
+      "step": 613
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 614
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 615
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.6695,
+      "step": 616
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.3909,
+      "step": 617
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 618
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.2673,
+      "step": 619
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 620
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.2726,
+      "step": 621
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.213,
+      "step": 622
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.2516,
+      "step": 623
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 624
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 2.104,
+      "step": 625
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.9442,
+      "step": 626
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 627
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 628
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.9279,
+      "step": 629
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 630
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 631
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 632
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.4213,
+      "step": 633
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 634
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 635
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 2.0403,
+      "step": 636
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 637
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.2558,
+      "step": 638
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.1418,
+      "step": 639
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.0833,
+      "step": 640
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.8653,
+      "step": 641
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.2917,
+      "step": 642
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 643
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.3209,
+      "step": 644
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.2496,
+      "step": 645
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 646
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 647
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 2.1891,
+      "step": 648
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 649
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 650
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.7846,
+      "step": 651
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.2405,
+      "step": 652
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.2027,
+      "step": 653
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.334,
+      "step": 654
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 2.3125,
+      "step": 655
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 656
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.4278,
+      "step": 657
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 658
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 659
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.763,
+      "step": 660
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 661
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 662
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 663
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.3284,
+      "step": 664
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 665
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.3993,
+      "step": 666
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 667
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001,
+      "loss": 1.3494,
+      "step": 668
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.3654,
+      "step": 669
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 670
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 671
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.238,
+      "step": 672
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.8114,
+      "step": 673
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 674
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 675
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.1257,
+      "step": 676
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 677
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 2.072,
+      "step": 678
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 679
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.8687,
+      "step": 680
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 681
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 682
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.7797,
+      "step": 683
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 684
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.7912,
+      "step": 685
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.3995,
+      "step": 686
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 687
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.4625,
+      "step": 688
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.2679,
+      "step": 689
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.9313,
+      "step": 690
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.2861,
+      "step": 691
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 692
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001,
+      "loss": 1.3878,
+      "step": 693
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 2.0583,
+      "step": 694
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 695
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 696
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.4482,
+      "step": 697
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.3106,
+      "step": 698
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 0.9151,
+      "step": 699
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.9587,
+      "step": 700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 701
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.8856,
+      "step": 702
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.9053,
+      "step": 703
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 2.1029,
+      "step": 704
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.3596,
+      "step": 705
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 2.054,
+      "step": 706
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.4661,
+      "step": 707
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.3528,
+      "step": 708
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 709
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 2.1202,
+      "step": 710
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 711
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.36,
+      "step": 712
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 713
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 714
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 2.0261,
+      "step": 715
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 716
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.364,
+      "step": 717
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.3405,
+      "step": 718
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 719
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.9452,
+      "step": 720
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 721
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 722
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 2.0223,
+      "step": 723
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.9493,
+      "step": 724
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.781,
+      "step": 725
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 4.1688,
+      "step": 726
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.2743,
+      "step": 727
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 2.2043,
+      "step": 728
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 729
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 730
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 731
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 2.196,
+      "step": 732
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 733
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 734
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.5549,
+      "step": 735
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.8167,
+      "step": 736
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 737
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.2442,
+      "step": 738
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 739
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.8608,
+      "step": 740
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 741
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 742
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.7371,
+      "step": 743
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.1784,
+      "step": 744
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.781,
+      "step": 745
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.7604,
+      "step": 746
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 747
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.8939,
+      "step": 748
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 749
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.0275,
+      "step": 750
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 751
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.8563,
+      "step": 752
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.9067,
+      "step": 753
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 754
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.8553,
+      "step": 755
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.1167,
+      "step": 756
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 2.0067,
+      "step": 757
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 758
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.1431,
+      "step": 759
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 760
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.7483,
+      "step": 761
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.7668,
+      "step": 762
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 763
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.9153,
+      "step": 764
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 765
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 766
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 767
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 768
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 769
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.3896,
+      "step": 770
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.8272,
+      "step": 771
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 772
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.6982,
+      "step": 773
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 774
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 775
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 776
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.2917,
+      "step": 777
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 778
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.7029,
+      "step": 779
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 780
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 2.2367,
+      "step": 781
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.8318,
+      "step": 782
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.9015,
+      "step": 783
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 784
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.3554,
+      "step": 785
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 786
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 787
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 788
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 789
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.3617,
+      "step": 790
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.4051,
+      "step": 791
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.7932,
+      "step": 792
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 793
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 794
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.8684,
+      "step": 795
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 796
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.0872,
+      "step": 797
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 798
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.72,
+      "step": 799
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.9553,
+      "step": 800
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 801
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 802
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.8419,
+      "step": 803
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.7891,
+      "step": 804
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.8436,
+      "step": 805
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 806
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.3858,
+      "step": 807
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.9157,
+      "step": 808
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.3744,
+      "step": 809
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 810
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.5228,
+      "step": 811
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.0831,
+      "step": 812
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 813
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 814
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 815
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.3216,
+      "step": 816
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.9641,
+      "step": 817
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.7587,
+      "step": 818
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001,
+      "loss": 1.9704,
+      "step": 819
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.3356,
+      "step": 820
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 821
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.8401,
+      "step": 822
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.2967,
+      "step": 823
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 824
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.7445,
+      "step": 825
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.1293,
+      "step": 826
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 827
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.3246,
+      "step": 828
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.0037,
+      "step": 829
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 830
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.1759,
+      "step": 831
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.9073,
+      "step": 832
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.9243,
+      "step": 833
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 834
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.2151,
+      "step": 835
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 836
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 837
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.2944,
+      "step": 838
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.226,
+      "step": 839
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 840
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.4425,
+      "step": 841
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.4395,
+      "step": 842
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 843
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 844
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 845
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.8849,
+      "step": 846
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 847
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.4272,
+      "step": 848
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.7245,
+      "step": 849
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.3521,
+      "step": 850
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 851
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 852
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.1811,
+      "step": 853
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.342,
+      "step": 854
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.7595,
+      "step": 855
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 856
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 857
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 858
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.2325,
+      "step": 859
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 860
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 861
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 862
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 863
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.3645,
+      "step": 864
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 865
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 866
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 867
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 868
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 1.3695,
+      "step": 869
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001,
+      "loss": 2.0724,
+      "step": 870
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 871
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 872
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.7746,
+      "step": 873
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.8178,
+      "step": 874
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 875
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.7716,
+      "step": 876
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 877
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 878
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 879
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 880
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.2137,
+      "step": 881
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 2.1077,
+      "step": 882
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 883
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 884
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 885
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.7751,
+      "step": 886
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 887
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.3123,
+      "step": 888
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.2515,
+      "step": 889
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 890
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.4038,
+      "step": 891
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.7371,
+      "step": 892
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.7505,
+      "step": 893
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 894
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001,
+      "loss": 1.2799,
+      "step": 895
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 896
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.3927,
+      "step": 897
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.6109,
+      "step": 898
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.1538,
+      "step": 899
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 900
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.9271,
+      "step": 901
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 902
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.3795,
+      "step": 903
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 904
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 905
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 906
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.3273,
+      "step": 907
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.7508,
+      "step": 908
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.4237,
+      "step": 909
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.3522,
+      "step": 910
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.7091,
+      "step": 911
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 912
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 913
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 914
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 915
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.9184,
+      "step": 916
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 917
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.4957,
+      "step": 918
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 919
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001,
+      "loss": 1.7357,
+      "step": 920
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.3281,
+      "step": 921
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 922
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.2176,
+      "step": 923
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.2323,
+      "step": 924
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.2509,
+      "step": 925
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.7233,
+      "step": 926
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 927
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 928
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 929
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 930
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.8867,
+      "step": 931
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 932
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 933
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 934
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 935
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 936
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.9009,
+      "step": 937
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 938
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.3152,
+      "step": 939
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.8733,
+      "step": 940
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.7578,
+      "step": 941
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 942
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 2.0986,
+      "step": 943
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 944
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 945
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 946
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.8096,
+      "step": 947
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 948
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.3472,
+      "step": 949
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.5965,
+      "step": 950
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.7525,
+      "step": 951
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 952
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 953
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.7068,
+      "step": 954
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 955
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.1832,
+      "step": 956
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.9865,
+      "step": 957
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.705,
+      "step": 958
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.3709,
+      "step": 959
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.8848,
+      "step": 960
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.8232,
+      "step": 961
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 962
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.923,
+      "step": 963
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 964
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 965
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 966
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.5676,
+      "step": 967
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.9132,
+      "step": 968
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.7008,
+      "step": 969
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.3316,
+      "step": 970
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 971
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.8195,
+      "step": 972
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 973
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.2923,
+      "step": 974
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 975
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.3759,
+      "step": 976
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.3873,
+      "step": 977
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 2.147,
+      "step": 978
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 979
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.9121,
+      "step": 980
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.3437,
+      "step": 981
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.8537,
+      "step": 982
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 983
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 984
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 985
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.4356,
+      "step": 986
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.2276,
+      "step": 987
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 988
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 989
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.3531,
+      "step": 990
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 2.142,
+      "step": 991
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.376,
+      "step": 992
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.3833,
+      "step": 993
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 994
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.3422,
+      "step": 995
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 996
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 997
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 998
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 999
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.3929,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.3016,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.2944,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.7054,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.8329,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.3513,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.9445,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.3275,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.875,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.8153,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.3906,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001,
+      "loss": 1.1968,
+      "step": 1021
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 1022
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 1023
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.9782,
+      "step": 1024
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 1025
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.823,
+      "step": 1026
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 1027
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.4634,
+      "step": 1028
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.8711,
+      "step": 1029
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.2189,
+      "step": 1030
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.7696,
+      "step": 1031
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.3714,
+      "step": 1032
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 1033
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.8065,
+      "step": 1034
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 1035
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.1857,
+      "step": 1036
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.1933,
+      "step": 1037
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 1038
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.8915,
+      "step": 1039
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 1040
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 1041
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.92,
+      "step": 1042
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.3125,
+      "step": 1043
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.681,
+      "step": 1044
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.03,
+      "step": 1045
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 1.9179,
+      "step": 1046
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001,
+      "loss": 2.0256,
+      "step": 1047
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.017,
+      "step": 1048
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.0832,
+      "step": 1049
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.3667,
+      "step": 1050
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.9458,
+      "step": 1051
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 1052
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.3603,
+      "step": 1053
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 1054
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.3926,
+      "step": 1055
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 1056
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.8837,
+      "step": 1057
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.8503,
+      "step": 1058
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 1059
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.345,
+      "step": 1060
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.8834,
+      "step": 1061
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 1062
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.5216,
+      "step": 1063
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.3077,
+      "step": 1064
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 1065
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.3756,
+      "step": 1066
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 1067
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 1068
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.1835,
+      "step": 1069
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 2.0085,
+      "step": 1070
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.3081,
+      "step": 1071
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 1072
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.7916,
+      "step": 1073
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.2537,
+      "step": 1074
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.4223,
+      "step": 1075
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.3501,
+      "step": 1076
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.4079,
+      "step": 1077
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 1078
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.1215,
+      "step": 1079
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.8549,
+      "step": 1080
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.7652,
+      "step": 1081
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.9078,
+      "step": 1082
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.289,
+      "step": 1083
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.7572,
+      "step": 1084
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.9276,
+      "step": 1085
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.4176,
+      "step": 1086
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.4136,
+      "step": 1087
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 1088
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.7399,
+      "step": 1089
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 1090
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 1091
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 1092
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.4304,
+      "step": 1093
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.8659,
+      "step": 1094
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 2.0272,
+      "step": 1095
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.896,
+      "step": 1096
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001,
+      "loss": 1.2737,
+      "step": 1097
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.2075,
+      "step": 1098
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.7437,
+      "step": 1099
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.8404,
+      "step": 1100
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 1101
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 1102
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 2.0779,
+      "step": 1103
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.0387,
+      "step": 1104
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.3666,
+      "step": 1105
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.354,
+      "step": 1106
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 1107
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 0.9545,
+      "step": 1108
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 1109
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.6168,
+      "step": 1110
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.1996,
+      "step": 1111
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.9961,
+      "step": 1112
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.7996,
+      "step": 1113
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 2.016,
+      "step": 1114
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.3233,
+      "step": 1115
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 2.0198,
+      "step": 1116
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.7238,
+      "step": 1117
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.3957,
+      "step": 1118
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 1119
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.3512,
+      "step": 1120
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 1121
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 1122
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.4107,
+      "step": 1123
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 1124
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 1125
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.8037,
+      "step": 1126
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.8449,
+      "step": 1127
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 1128
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.9726,
+      "step": 1129
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 1130
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 1131
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 1132
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.842,
+      "step": 1133
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.4116,
+      "step": 1134
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.3831,
+      "step": 1135
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 2.018,
+      "step": 1136
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 1137
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 1138
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.64,
+      "step": 1139
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 1140
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.3739,
+      "step": 1141
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.4452,
+      "step": 1142
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 1143
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.3416,
+      "step": 1144
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.3328,
+      "step": 1145
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 1146
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001,
+      "loss": 1.2555,
+      "step": 1147
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.4256,
+      "step": 1148
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 1149
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.9,
+      "step": 1150
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 1151
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 1152
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 1153
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 1154
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.3419,
+      "step": 1155
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.3474,
+      "step": 1156
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 1157
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.7387,
+      "step": 1158
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.9022,
+      "step": 1159
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 1160
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 1161
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.7926,
+      "step": 1162
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.7802,
+      "step": 1163
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.285,
+      "step": 1164
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 1165
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 1166
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 1167
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 1168
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 1169
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.1945,
+      "step": 1170
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.318,
+      "step": 1171
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 1172
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 1173
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.3239,
+      "step": 1174
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 1175
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.2919,
+      "step": 1176
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 2.0054,
+      "step": 1177
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 1178
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 1179
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.4702,
+      "step": 1180
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 1181
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 1182
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.4538,
+      "step": 1183
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.7568,
+      "step": 1184
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 1185
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 1186
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.7284,
+      "step": 1187
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 1188
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.9413,
+      "step": 1189
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 1190
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.8242,
+      "step": 1191
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.1322,
+      "step": 1192
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 1193
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.0586,
+      "step": 1194
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.1677,
+      "step": 1195
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.494,
+      "step": 1196
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 1197
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 1198
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 1199
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.8501,
+      "step": 1200
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.3564,
+      "step": 1201
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.8763,
+      "step": 1202
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.3294,
+      "step": 1203
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 1204
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.3421,
+      "step": 1205
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 0.9211,
+      "step": 1206
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 1207
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 1208
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 2.1296,
+      "step": 1209
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.8805,
+      "step": 1210
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.4274,
+      "step": 1211
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.2607,
+      "step": 1212
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 1213
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.7839,
+      "step": 1214
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 1215
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 1216
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.7227,
+      "step": 1217
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 1218
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 1219
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 1220
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.2347,
+      "step": 1221
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.9851,
+      "step": 1222
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001,
+      "loss": 1.2311,
+      "step": 1223
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 1224
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.7614,
+      "step": 1225
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 1226
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 1227
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.8767,
+      "step": 1228
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 1229
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 1230
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 1231
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.2225,
+      "step": 1232
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 1233
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 1234
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 1235
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 1236
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 1237
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.8736,
+      "step": 1238
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.1765,
+      "step": 1239
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.8874,
+      "step": 1240
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.3572,
+      "step": 1241
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 1242
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 1243
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.9905,
+      "step": 1244
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 1245
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 1246
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.3952,
+      "step": 1247
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001,
+      "loss": 1.9669,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.8839,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.7578,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.0617,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.3755,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.3837,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.1495,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.8659,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.3348,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.3449,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.895,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.9958,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.7767,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.1731,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 1.9365,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001,
+      "loss": 2.1234,
+      "step": 1274
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 1275
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 1276
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 2.2804,
+      "step": 1277
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 1278
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.7985,
+      "step": 1279
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 2.04,
+      "step": 1280
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 1281
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.8126,
+      "step": 1282
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.8663,
+      "step": 1283
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 2.0001,
+      "step": 1284
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 1285
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.9925,
+      "step": 1286
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.4277,
+      "step": 1287
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 1288
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 1289
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 1290
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.2909,
+      "step": 1291
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.4391,
+      "step": 1292
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 1293
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 1294
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.9872,
+      "step": 1295
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.5225,
+      "step": 1296
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.7299,
+      "step": 1297
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 1298
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001,
+      "loss": 1.8119,
+      "step": 1299
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.2642,
+      "step": 1300
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.3914,
+      "step": 1301
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.0188,
+      "step": 1302
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 1303
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 1304
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.7096,
+      "step": 1305
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.3454,
+      "step": 1306
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.3258,
+      "step": 1307
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 1308
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.7414,
+      "step": 1309
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.0357,
+      "step": 1310
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 1311
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 1312
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 1313
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 1314
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.5254,
+      "step": 1315
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.1985,
+      "step": 1316
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 1317
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.8928,
+      "step": 1318
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 1319
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 1320
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.2643,
+      "step": 1321
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.844,
+      "step": 1322
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.557,
+      "step": 1323
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 1324
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 2.022,
+      "step": 1325
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 1326
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 1327
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 1328
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 1329
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.8739,
+      "step": 1330
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.4369,
+      "step": 1331
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 1332
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 1333
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.6021,
+      "step": 1334
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 2.0535,
+      "step": 1335
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 1336
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 1337
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.9306,
+      "step": 1338
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.8141,
+      "step": 1339
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 1340
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.2051,
+      "step": 1341
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 2.1542,
+      "step": 1342
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.8667,
+      "step": 1343
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 1344
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.3398,
+      "step": 1345
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 1346
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 1347
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 1348
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001,
+      "loss": 1.7827,
+      "step": 1349
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4161,
+      "step": 1350
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4278,
+      "step": 1351
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 1352
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.7204,
+      "step": 1353
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.8292,
+      "step": 1354
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.221,
+      "step": 1355
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 1356
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 1357
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 1358
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.8561,
+      "step": 1359
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 2.0282,
+      "step": 1360
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.3186,
+      "step": 1361
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 1362
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 2.0099,
+      "step": 1363
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 1364
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.1815,
+      "step": 1365
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 1366
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.2678,
+      "step": 1367
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 1368
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.3218,
+      "step": 1369
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.7572,
+      "step": 1370
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 1371
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.5252,
+      "step": 1372
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 1373
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.4382,
+      "step": 1374
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 1375
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 1376
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 1377
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 1378
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 1379
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 1380
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 2.0154,
+      "step": 1381
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 1382
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 1383
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.6883,
+      "step": 1384
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.82,
+      "step": 1385
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.892,
+      "step": 1386
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 1387
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.975,
+      "step": 1388
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 1389
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 1390
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.1982,
+      "step": 1391
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.4121,
+      "step": 1392
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 1393
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.7656,
+      "step": 1394
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.8195,
+      "step": 1395
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.2235,
+      "step": 1396
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 1397
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 1398
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 1399
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001,
+      "loss": 1.2293,
+      "step": 1400
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 1401
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 1402
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.1955,
+      "step": 1403
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 1404
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.2547,
+      "step": 1405
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 1406
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.3022,
+      "step": 1407
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 1408
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 1409
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 1410
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.8193,
+      "step": 1411
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 2.1626,
+      "step": 1412
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 1413
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 1414
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.7777,
+      "step": 1415
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.9457,
+      "step": 1416
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 1417
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 1418
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.2212,
+      "step": 1419
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.4908,
+      "step": 1420
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 1421
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 1422
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 1423
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 1424
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 1425
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.3887,
+      "step": 1426
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.2372,
+      "step": 1427
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 1428
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.8664,
+      "step": 1429
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 1430
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.9392,
+      "step": 1431
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.4137,
+      "step": 1432
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 1433
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.784,
+      "step": 1434
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.8056,
+      "step": 1435
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.8309,
+      "step": 1436
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 1437
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.2986,
+      "step": 1438
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.3181,
+      "step": 1439
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 1440
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.4573,
+      "step": 1441
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 2.2395,
+      "step": 1442
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 1443
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 1444
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.8278,
+      "step": 1445
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.8702,
+      "step": 1446
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.3345,
+      "step": 1447
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.8964,
+      "step": 1448
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.4353,
+      "step": 1449
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001,
+      "loss": 1.3406,
+      "step": 1450
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.894,
+      "step": 1451
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 1452
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 1453
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 1454
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 1455
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.8619,
+      "step": 1456
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.8269,
+      "step": 1457
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 1458
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.3425,
+      "step": 1459
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 1460
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 1461
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.4378,
+      "step": 1462
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.3498,
+      "step": 1463
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 1464
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 1465
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.3604,
+      "step": 1466
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 1467
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.8037,
+      "step": 1468
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 2.0071,
+      "step": 1469
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 1470
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.3872,
+      "step": 1471
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 1472
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 1473
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 1474
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001,
+      "loss": 1.9373,
+      "step": 1475
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.3856,
+      "step": 1476
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.1911,
+      "step": 1477
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 1478
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 1479
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.9612,
+      "step": 1480
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 1481
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.2123,
+      "step": 1482
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.8292,
+      "step": 1483
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.3318,
+      "step": 1484
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 1485
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.7052,
+      "step": 1486
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 1487
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 1488
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.7388,
+      "step": 1489
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 1490
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 1491
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 2.1285,
+      "step": 1492
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 1493
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 1494
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 1495
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.2966,
+      "step": 1496
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.3919,
+      "step": 1497
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.0465,
+      "step": 1498
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 1499
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.3862,
+      "step": 1500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.8062,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.6325,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.2666,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.3116,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.1723,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.9469,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.221,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.9039,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.8432,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.3045,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.9642,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.2066,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.2224,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.348,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.3066,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 1526
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 1527
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.351,
+      "step": 1528
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 1529
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 1530
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.3239,
+      "step": 1531
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.8906,
+      "step": 1532
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.0036,
+      "step": 1533
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.8105,
+      "step": 1534
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.7316,
+      "step": 1535
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.9053,
+      "step": 1536
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 1537
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 1538
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.9442,
+      "step": 1539
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 1540
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 1541
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.7171,
+      "step": 1542
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.2705,
+      "step": 1543
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 1544
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.2349,
+      "step": 1545
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 1546
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 1547
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.1645,
+      "step": 1548
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.3955,
+      "step": 1549
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 1550
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 1551
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 1552
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 2.0603,
+      "step": 1553
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 1554
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.4142,
+      "step": 1555
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.4471,
+      "step": 1556
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.8487,
+      "step": 1557
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.4116,
+      "step": 1558
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 1559
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.836,
+      "step": 1560
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.6623,
+      "step": 1561
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 1562
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 1563
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 1564
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 1565
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 1566
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.178,
+      "step": 1567
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.0348,
+      "step": 1568
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.2936,
+      "step": 1569
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 1570
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 1571
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 1572
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.3649,
+      "step": 1573
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 1574
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.4131,
+      "step": 1575
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001,
+      "loss": 1.9998,
+      "step": 1576
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 1577
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 2.2141,
+      "step": 1578
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 2.1626,
+      "step": 1579
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 1580
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.9964,
+      "step": 1581
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 1582
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 2.1103,
+      "step": 1583
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 1584
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 1585
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.4279,
+      "step": 1586
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.9997,
+      "step": 1587
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 2.0111,
+      "step": 1588
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 1589
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.3203,
+      "step": 1590
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 1591
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.3977,
+      "step": 1592
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 1593
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.3901,
+      "step": 1594
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.3519,
+      "step": 1595
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 1596
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 1597
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 1598
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 2.0256,
+      "step": 1599
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.4594,
+      "step": 1600
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.7696,
+      "step": 1601
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 1602
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.7101,
+      "step": 1603
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 1604
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 1605
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 1606
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 1607
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 1608
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 1609
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 1610
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 1611
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 1612
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 1613
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 1614
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.7952,
+      "step": 1615
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 1616
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 1617
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 1618
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 2.1767,
+      "step": 1619
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.3179,
+      "step": 1620
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.1073,
+      "step": 1621
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 1622
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.4785,
+      "step": 1623
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 1624
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 1625
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.2956,
+      "step": 1626
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001,
+      "loss": 1.8781,
+      "step": 1627
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.9721,
+      "step": 1628
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.8693,
+      "step": 1629
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 1630
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.7886,
+      "step": 1631
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 1632
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.916,
+      "step": 1633
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.8049,
+      "step": 1634
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.7312,
+      "step": 1635
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 2.0517,
+      "step": 1636
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 1637
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 1638
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.8448,
+      "step": 1639
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.7502,
+      "step": 1640
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 1641
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.848,
+      "step": 1642
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.3757,
+      "step": 1643
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.1445,
+      "step": 1644
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 1645
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.354,
+      "step": 1646
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 2.2129,
+      "step": 1647
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 1648
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.873,
+      "step": 1649
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.8591,
+      "step": 1650
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.7656,
+      "step": 1651
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001,
+      "loss": 1.1412,
+      "step": 1652
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.214,
+      "step": 1653
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.6908,
+      "step": 1654
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 2.1402,
+      "step": 1655
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 1656
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 1657
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 1658
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.4756,
+      "step": 1659
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 1660
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.8135,
+      "step": 1661
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.8093,
+      "step": 1662
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.9819,
+      "step": 1663
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 1664
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 1665
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.7951,
+      "step": 1666
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 1667
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 1668
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 1669
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 1670
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.3344,
+      "step": 1671
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 1672
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 1673
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.849,
+      "step": 1674
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 1675
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 1.2024,
+      "step": 1676
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001,
+      "loss": 2.0003,
+      "step": 1677
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.2433,
+      "step": 1678
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 1679
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 1680
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.937,
+      "step": 1681
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 1682
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 1683
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 1684
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 1685
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.0923,
+      "step": 1686
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 1687
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.2753,
+      "step": 1688
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 1689
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.3789,
+      "step": 1690
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 1691
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 1692
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.9369,
+      "step": 1693
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.3422,
+      "step": 1694
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.3955,
+      "step": 1695
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.8257,
+      "step": 1696
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 1697
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 1698
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.8539,
+      "step": 1699
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.2594,
+      "step": 1700
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.2933,
+      "step": 1701
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 1702
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001,
+      "loss": 1.8121,
+      "step": 1703
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.784,
+      "step": 1704
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.3501,
+      "step": 1705
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 1706
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 1707
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.1769,
+      "step": 1708
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.3047,
+      "step": 1709
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.355,
+      "step": 1710
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 1711
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 1712
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 2.165,
+      "step": 1713
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 1714
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 1715
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 1716
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.9517,
+      "step": 1717
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.8701,
+      "step": 1718
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.9787,
+      "step": 1719
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 1720
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 1721
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 1722
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.5896,
+      "step": 1723
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.2121,
+      "step": 1724
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 1725
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 1726
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.8909,
+      "step": 1727
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 1728
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 1729
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 1730
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.8341,
+      "step": 1731
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.1973,
+      "step": 1732
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 1733
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.4477,
+      "step": 1734
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.2387,
+      "step": 1735
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 1736
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 1737
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.7478,
+      "step": 1738
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 1739
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.2412,
+      "step": 1740
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 1741
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 1742
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.2806,
+      "step": 1743
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 1744
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 1745
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.349,
+      "step": 1746
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 1747
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 1748
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 1749
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.3462,
+      "step": 1750
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 1751
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 1752
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001,
+      "loss": 1.8965,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.4255,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.8803,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.8255,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.4237,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.7376,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.1254,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.2906,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.4163,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.7994,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.8642,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 1.2791,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001,
+      "loss": 2.3644,
+      "step": 1778
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.3548,
+      "step": 1779
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.2818,
+      "step": 1780
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 1781
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.4027,
+      "step": 1782
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.3384,
+      "step": 1783
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 1784
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.4383,
+      "step": 1785
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 1786
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.2273,
+      "step": 1787
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.3816,
+      "step": 1788
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 1789
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 1790
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.3193,
+      "step": 1791
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.889,
+      "step": 1792
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.1186,
+      "step": 1793
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.4801,
+      "step": 1794
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.119,
+      "step": 1795
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 1796
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.9109,
+      "step": 1797
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.3755,
+      "step": 1798
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 1799
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.2717,
+      "step": 1800
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.3574,
+      "step": 1801
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.2837,
+      "step": 1802
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 1803
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 1804
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.3289,
+      "step": 1805
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.1418,
+      "step": 1806
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.2667,
+      "step": 1807
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.6804,
+      "step": 1808
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.2312,
+      "step": 1809
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.825,
+      "step": 1810
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 1811
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.3245,
+      "step": 1812
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.3535,
+      "step": 1813
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.2794,
+      "step": 1814
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 1815
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.2967,
+      "step": 1816
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 1817
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.2223,
+      "step": 1818
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 1819
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 1820
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.3876,
+      "step": 1821
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 1822
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 1823
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 1824
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.2754,
+      "step": 1825
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 1826
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.9491,
+      "step": 1827
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 2.0952,
+      "step": 1828
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001,
+      "loss": 1.3997,
+      "step": 1829
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.8559,
+      "step": 1830
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 1831
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.9345,
+      "step": 1832
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 1833
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.428,
+      "step": 1834
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 0.8477,
+      "step": 1835
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.3121,
+      "step": 1836
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.4888,
+      "step": 1837
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.2423,
+      "step": 1838
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 1839
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 1840
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 1841
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 1842
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.9425,
+      "step": 1843
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 1844
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.8733,
+      "step": 1845
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 1846
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 1847
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 1848
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 2.0855,
+      "step": 1849
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 2.2499,
+      "step": 1850
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 1851
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.3094,
+      "step": 1852
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.4801,
+      "step": 1853
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 1854
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.1657,
+      "step": 1855
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 1856
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.2637,
+      "step": 1857
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.2036,
+      "step": 1858
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.1982,
+      "step": 1859
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.9378,
+      "step": 1860
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 1861
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 1862
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 1863
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.2688,
+      "step": 1864
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 1865
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.7747,
+      "step": 1866
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 1867
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.2336,
+      "step": 1868
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.4416,
+      "step": 1869
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 1870
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.335,
+      "step": 1871
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.4351,
+      "step": 1872
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 1873
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 1874
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 1875
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 1876
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 2.0429,
+      "step": 1877
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.7444,
+      "step": 1878
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 1879
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.8131,
+      "step": 1880
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 1881
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 1882
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 1883
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.8837,
+      "step": 1884
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 2.0464,
+      "step": 1885
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 1886
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 1887
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 1888
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 1889
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 1890
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.1844,
+      "step": 1891
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 1892
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 1893
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.1667,
+      "step": 1894
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.0057,
+      "step": 1895
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 1896
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 1897
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 1898
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 1899
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 1900
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 1901
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.4863,
+      "step": 1902
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 1903
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 1904
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.9988,
+      "step": 1905
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.915,
+      "step": 1906
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 1907
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.849,
+      "step": 1908
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.3067,
+      "step": 1909
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.803,
+      "step": 1910
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.3386,
+      "step": 1911
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 1912
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 1913
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.2556,
+      "step": 1914
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1915
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.3434,
+      "step": 1916
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.2375,
+      "step": 1917
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.9422,
+      "step": 1918
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.2539,
+      "step": 1919
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.051,
+      "step": 1920
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.3183,
+      "step": 1921
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 2.1107,
+      "step": 1922
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 1923
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.5024,
+      "step": 1924
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 1925
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.2468,
+      "step": 1926
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.2835,
+      "step": 1927
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.419,
+      "step": 1928
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.3431,
+      "step": 1929
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001,
+      "loss": 1.3103,
+      "step": 1930
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 1931
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 1932
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.7988,
+      "step": 1933
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.1145,
+      "step": 1934
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 1935
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.3721,
+      "step": 1936
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 1937
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.2809,
+      "step": 1938
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 1939
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 1940
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.8963,
+      "step": 1941
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 2.0296,
+      "step": 1942
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 1943
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 1944
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 1945
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 1946
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.8559,
+      "step": 1947
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 2.0111,
+      "step": 1948
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 1949
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 1950
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 1951
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 1952
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 1953
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 1.3751,
+      "step": 1954
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001,
+      "loss": 2.1813,
+      "step": 1955
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 1956
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 1957
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.9329,
+      "step": 1958
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.8971,
+      "step": 1959
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.3074,
+      "step": 1960
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.2736,
+      "step": 1961
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 1962
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.8975,
+      "step": 1963
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 1964
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 1965
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 1966
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.3303,
+      "step": 1967
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 1968
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.4014,
+      "step": 1969
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 1970
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.3006,
+      "step": 1971
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 1972
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.3296,
+      "step": 1973
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 1974
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 1975
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.9429,
+      "step": 1976
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 2.3806,
+      "step": 1977
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.9194,
+      "step": 1978
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.2388,
+      "step": 1979
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 1980
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 1981
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.369,
+      "step": 1982
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 1983
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.265,
+      "step": 1984
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 1985
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 1986
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 1987
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 1988
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5257,
+      "step": 1989
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.2654,
+      "step": 1990
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.9902,
+      "step": 1991
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 1992
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.7872,
+      "step": 1993
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.6776,
+      "step": 1994
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 1995
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 1996
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 1997
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 1998
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 1999
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 2.0002,
+      "step": 2000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 2001
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 2.2816,
+      "step": 2002
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 2003
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.1632,
+      "step": 2004
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001,
+      "loss": 1.8702,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.8423,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 2.04,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.2414,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.8668,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 2015
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.2588,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.3056,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 2.0218,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.7191,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.1817,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.449,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 2.1054,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.3467,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.9854,
+      "step": 2027
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.2397,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 2.284,
+      "step": 2029
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.1385,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001,
+      "loss": 1.3635,
+      "step": 2031
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.346,
+      "step": 2032
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 2033
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.6696,
+      "step": 2034
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 2035
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.2051,
+      "step": 2036
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.7849,
+      "step": 2037
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 2038
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.1109,
+      "step": 2039
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 2040
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.9772,
+      "step": 2041
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.3313,
+      "step": 2042
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 2043
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.9321,
+      "step": 2044
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.9089,
+      "step": 2045
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 2046
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.9344,
+      "step": 2047
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 2048
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 2049
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 2050
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 2051
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.3417,
+      "step": 2052
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.3794,
+      "step": 2053
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 2054
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.7919,
+      "step": 2055
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 2056
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 2057
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.39,
+      "step": 2058
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 2059
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.7891,
+      "step": 2060
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.3487,
+      "step": 2061
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.3686,
+      "step": 2062
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.9134,
+      "step": 2063
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 2064
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.7842,
+      "step": 2065
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.9323,
+      "step": 2066
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.9835,
+      "step": 2067
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.2889,
+      "step": 2068
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 2069
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 0.9831,
+      "step": 2070
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.1389,
+      "step": 2071
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.2248,
+      "step": 2072
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.3999,
+      "step": 2073
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.1983,
+      "step": 2074
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.2714,
+      "step": 2075
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 2.0523,
+      "step": 2076
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 2077
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 2078
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.2981,
+      "step": 2079
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.7939,
+      "step": 2080
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001,
+      "loss": 1.7438,
+      "step": 2081
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.1959,
+      "step": 2082
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.7568,
+      "step": 2083
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 2084
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 2085
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.7606,
+      "step": 2086
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.8548,
+      "step": 2087
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 2088
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.1236,
+      "step": 2089
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 2090
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 2.0006,
+      "step": 2091
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 2092
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.3562,
+      "step": 2093
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.1095,
+      "step": 2094
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.3963,
+      "step": 2095
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.2097,
+      "step": 2096
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 2097
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.3236,
+      "step": 2098
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.1953,
+      "step": 2099
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.3048,
+      "step": 2100
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 2101
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.8415,
+      "step": 2102
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.2222,
+      "step": 2103
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 2104
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.3723,
+      "step": 2105
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 2106
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.2131,
+      "step": 2107
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 2108
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.3369,
+      "step": 2109
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 2.2784,
+      "step": 2110
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.1495,
+      "step": 2111
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 2112
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.7329,
+      "step": 2113
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.3691,
+      "step": 2114
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.172,
+      "step": 2115
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.3816,
+      "step": 2116
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 2117
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.9691,
+      "step": 2118
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.3523,
+      "step": 2119
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 2120
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 2121
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.9666,
+      "step": 2122
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 2123
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 2124
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 2125
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 2126
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 2127
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 2128
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.4634,
+      "step": 2129
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.1596,
+      "step": 2130
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001,
+      "loss": 1.4321,
+      "step": 2131
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 2132
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.2369,
+      "step": 2133
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.5233,
+      "step": 2134
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.7539,
+      "step": 2135
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.3513,
+      "step": 2136
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 2137
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.9485,
+      "step": 2138
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 2139
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.8767,
+      "step": 2140
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 2141
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 2142
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.788,
+      "step": 2143
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.4336,
+      "step": 2144
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 2145
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.3072,
+      "step": 2146
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 2.1035,
+      "step": 2147
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 2148
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 2149
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 2150
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 2151
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 2152
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 2153
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 2154
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 2155
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 2.1172,
+      "step": 2156
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001,
+      "loss": 1.2135,
+      "step": 2157
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 2158
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 2159
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.4171,
+      "step": 2160
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.2318,
+      "step": 2161
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 2162
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.1614,
+      "step": 2163
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 2164
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 2165
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 2166
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.2386,
+      "step": 2167
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 2168
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 2169
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 2170
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 2171
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.7216,
+      "step": 2172
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.4524,
+      "step": 2173
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 2174
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 2175
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.1187,
+      "step": 2176
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 0.9964,
+      "step": 2177
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.3753,
+      "step": 2178
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.1551,
+      "step": 2179
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.0277,
+      "step": 2180
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.4002,
+      "step": 2181
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 2182
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.4556,
+      "step": 2183
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 2184
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 2185
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.283,
+      "step": 2186
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.9317,
+      "step": 2187
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 2188
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.3377,
+      "step": 2189
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 0.8071,
+      "step": 2190
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.3542,
+      "step": 2191
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 2.0683,
+      "step": 2192
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.1612,
+      "step": 2193
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.4494,
+      "step": 2194
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 2195
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 2196
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.0553,
+      "step": 2197
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 2198
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 2199
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 2200
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 2201
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 2.0343,
+      "step": 2202
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.9364,
+      "step": 2203
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 2204
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.8149,
+      "step": 2205
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 2206
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 2207
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.5163,
+      "step": 2208
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.8425,
+      "step": 2209
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.314,
+      "step": 2210
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.8352,
+      "step": 2211
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 2212
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.2999,
+      "step": 2213
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 2214
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 2215
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.9026,
+      "step": 2216
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 2217
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.1616,
+      "step": 2218
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 2219
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 2220
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 2221
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 2.0223,
+      "step": 2222
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 2223
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 2224
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.9687,
+      "step": 2225
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 2226
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.2869,
+      "step": 2227
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 2228
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.2205,
+      "step": 2229
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.3398,
+      "step": 2230
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.0162,
+      "step": 2231
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 0.9122,
+      "step": 2232
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.0024,
+      "step": 2233
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.3844,
+      "step": 2234
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 2235
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.7781,
+      "step": 2236
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 2237
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.8227,
+      "step": 2238
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.3432,
+      "step": 2239
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 2240
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 2241
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 2242
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 2.0983,
+      "step": 2243
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 2244
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 2245
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.3995,
+      "step": 2246
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 2247
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 2248
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 2249
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 2.0653,
+      "step": 2250
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 2251
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 2252
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.2593,
+      "step": 2253
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 2254
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 2255
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.8149,
+      "step": 2256
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 2257
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001,
+      "loss": 1.6931,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 2259
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 2.3291,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.7503,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 2263
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.4254,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 2265
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.3597,
+      "step": 2267
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.3853,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.9018,
+      "step": 2269
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.2199,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.8328,
+      "step": 2271
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.2172,
+      "step": 2273
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.1377,
+      "step": 2275
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.0367,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.3474,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 2279
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.8904,
+      "step": 2281
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 2.1754,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 2283
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.4002,
+      "step": 2284
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.4156,
+      "step": 2285
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 2286
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 2287
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 2288
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 2289
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.7785,
+      "step": 2290
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 2291
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.2589,
+      "step": 2292
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 2293
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.331,
+      "step": 2294
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.4125,
+      "step": 2295
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.4002,
+      "step": 2296
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 2297
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.1653,
+      "step": 2298
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 2.016,
+      "step": 2299
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 2300
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 2.1736,
+      "step": 2301
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.3189,
+      "step": 2302
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.3152,
+      "step": 2303
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 2304
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 2305
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.8074,
+      "step": 2306
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.9449,
+      "step": 2307
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001,
+      "loss": 1.7673,
+      "step": 2308
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.3689,
+      "step": 2309
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 2310
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.2582,
+      "step": 2311
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.6912,
+      "step": 2312
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 2313
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.3459,
+      "step": 2314
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 2315
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.4283,
+      "step": 2316
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.9134,
+      "step": 2317
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.3453,
+      "step": 2318
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.7666,
+      "step": 2319
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 2320
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.3164,
+      "step": 2321
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.4216,
+      "step": 2322
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.2846,
+      "step": 2323
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.3114,
+      "step": 2324
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 2325
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 2326
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.2153,
+      "step": 2327
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 2328
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.3574,
+      "step": 2329
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.2886,
+      "step": 2330
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 2331
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 2332
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 2333
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 2334
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 2335
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.1144,
+      "step": 2336
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 2337
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.0578,
+      "step": 2338
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 2339
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.4872,
+      "step": 2340
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 2341
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 2342
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 2343
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 2344
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 2345
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 2346
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.4151,
+      "step": 2347
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 2348
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 2349
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 2350
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 2.0262,
+      "step": 2351
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 2352
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.2275,
+      "step": 2353
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.3499,
+      "step": 2354
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 2355
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.9356,
+      "step": 2356
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 2357
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.7099,
+      "step": 2358
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 2359
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 2360
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.3338,
+      "step": 2361
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.796,
+      "step": 2362
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 2363
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 2364
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 2365
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 2366
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 2367
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 2368
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.526,
+      "step": 2369
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 2370
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.9401,
+      "step": 2371
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.9511,
+      "step": 2372
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 2373
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 2374
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 2375
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.1469,
+      "step": 2376
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.0687,
+      "step": 2377
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 2378
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.4661,
+      "step": 2379
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 2380
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 2381
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 2382
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 2383
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 2384
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.2041,
+      "step": 2385
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 2386
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 2387
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 2.0364,
+      "step": 2388
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.6022,
+      "step": 2389
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 2390
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.3797,
+      "step": 2391
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.382,
+      "step": 2392
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 2393
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 2394
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 0.8872,
+      "step": 2395
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.3878,
+      "step": 2396
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 2397
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 2.5213,
+      "step": 2398
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 0.8851,
+      "step": 2399
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.2568,
+      "step": 2400
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.2932,
+      "step": 2401
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.2669,
+      "step": 2402
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.2796,
+      "step": 2403
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.3372,
+      "step": 2404
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.9432,
+      "step": 2405
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.3657,
+      "step": 2406
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.1607,
+      "step": 2407
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.2322,
+      "step": 2408
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001,
+      "loss": 1.1187,
+      "step": 2409
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 2410
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.4957,
+      "step": 2411
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.2321,
+      "step": 2412
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 2413
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 2414
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.1276,
+      "step": 2415
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.9021,
+      "step": 2416
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 2417
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.8233,
+      "step": 2418
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 2419
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.7064,
+      "step": 2420
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 2421
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.8513,
+      "step": 2422
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 2423
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.5408,
+      "step": 2424
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 2425
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 2426
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 2427
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 2428
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 2429
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 2430
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 2431
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.7852,
+      "step": 2432
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.3227,
+      "step": 2433
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 2434
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 2435
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 2436
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.7512,
+      "step": 2437
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 2438
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 2439
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 2440
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.2992,
+      "step": 2441
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 2442
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.8919,
+      "step": 2443
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 2444
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.2048,
+      "step": 2445
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.3682,
+      "step": 2446
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 2447
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.8932,
+      "step": 2448
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 2449
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 2450
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 2451
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.1833,
+      "step": 2452
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4133,
+      "step": 2453
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4322,
+      "step": 2454
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.9597,
+      "step": 2455
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 2456
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 2457
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.3404,
+      "step": 2458
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 2459
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 2460
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.8394,
+      "step": 2461
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 2.285,
+      "step": 2462
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.3341,
+      "step": 2463
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 2464
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.928,
+      "step": 2465
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 2466
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 2467
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 2468
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.9782,
+      "step": 2469
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 2.1464,
+      "step": 2470
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 2.0863,
+      "step": 2471
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 2472
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 2473
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 2474
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.4006,
+      "step": 2475
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.2611,
+      "step": 2476
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.2637,
+      "step": 2477
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.2899,
+      "step": 2478
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 2479
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.7978,
+      "step": 2480
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.4014,
+      "step": 2481
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.9763,
+      "step": 2482
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.2566,
+      "step": 2483
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 2484
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001,
+      "loss": 1.0761,
+      "step": 2485
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.8328,
+      "step": 2486
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 2487
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.1851,
+      "step": 2488
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.2698,
+      "step": 2489
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 2490
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.2989,
+      "step": 2491
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 2492
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 2493
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.901,
+      "step": 2494
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 2.159,
+      "step": 2495
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.1142,
+      "step": 2496
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.8338,
+      "step": 2497
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 2498
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 2499
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 2500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.4636,
+      "step": 2501
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.8867,
+      "step": 2502
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.3087,
+      "step": 2503
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.2474,
+      "step": 2504
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 2505
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.1296,
+      "step": 2506
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 2507
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 2508
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.7885,
+      "step": 2509
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001,
+      "loss": 1.3689,
+      "step": 2510
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.7565,
+      "step": 2511
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 2513
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.3344,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.1828,
+      "step": 2515
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.0794,
+      "step": 2516
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 2517
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.4082,
+      "step": 2519
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 2521
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 2523
+    }
+  ],
+  "max_steps": 7569,
+  "num_train_epochs": 3,
+  "total_flos": 3.0726749734714737e+18,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/training_args.bin b/training_args.bin
new file mode 100644
index 0000000..6365c98
--- /dev/null
+++ b/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdca217be1502b8b802b734f48aecd1555efbe7c0823c929d9a2a005cd9bbd10
+size 5307
diff --git a/zero_to_fp32.py b/zero_to_fp32.py
new file mode 100755
index 0000000..c98caae
--- /dev/null
+++ b/zero_to_fp32.py
@@ -0,0 +1,587 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel()
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, args.output_file, tag=args.tag)