From 25e40e38e87445959ad0af24069b745e1b6540fa Mon Sep 17 00:00:00 2001
From: ModelHub XC <noreply@modelhub.org.cn>
Date: Tue, 14 Jul 2026 20:14:11 +0800
Subject: [PATCH] =?UTF-8?q?=E5=88=9D=E5=A7=8B=E5=8C=96=E9=A1=B9=E7=9B=AE?=
 =?UTF-8?q?=EF=BC=8C=E7=94=B1ModelHub=20XC=E7=A4=BE=E5=8C=BA=E6=8F=90?=
 =?UTF-8?q?=E4=BE=9B=E6=A8=A1=E5=9E=8B?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Model: chenyongxi/Qwen2-1.5B-SFT-IF
Source: Original Platform
---
 .gitattributes         |   36 +
 README.md              |   57 ++
 chat_template.jinja    |   54 ++
 config.json            |   62 ++
 generation_config.json |   10 +
 model.safetensors      |    3 +
 rng_state.pth          |    3 +
 scheduler.pt           |    3 +
 tokenizer.json         |    3 +
 tokenizer_config.json  |   29 +
 trainer_state.json     | 1721 ++++++++++++++++++++++++++++++++++++++++
 training_args.bin      |    3 +
 12 files changed, 1984 insertions(+)
 create mode 100644 .gitattributes
 create mode 100644 README.md
 create mode 100644 chat_template.jinja
 create mode 100644 config.json
 create mode 100644 generation_config.json
 create mode 100644 model.safetensors
 create mode 100644 rng_state.pth
 create mode 100644 scheduler.pt
 create mode 100644 tokenizer.json
 create mode 100644 tokenizer_config.json
 create mode 100644 trainer_state.json
 create mode 100644 training_args.bin

diff --git a/.gitattributes b/.gitattributes
new file mode 100644
index 0000000..52373fe
--- /dev/null
+++ b/.gitattributes
@@ -0,0 +1,36 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt 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
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz 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
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl 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
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* 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
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..9b61a46
--- /dev/null
+++ b/README.md
@@ -0,0 +1,57 @@
+---
+base_model: Qwen/Qwen2.5-1.5B
+library_name: transformers
+model_name: Qwen2-1.5B-SFT-IF
+tags:
+- generated_from_trainer
+- trl
+- sft
+licence: license
+---
+
+# Model Card for Qwen2-1.5B-SFT-IF
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B](https://huggingface.co/Qwen/Qwen2.5-1.5B).
+It has been trained using [TRL](https://github.com/huggingface/trl).
+
+## Quick start
+
+```python
+from transformers import pipeline
+
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="chenyongxi/Qwen2-1.5B-SFT-IF", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/yongxichen83-cyx-test/huggingface/runs/1jejbszt) 
+
+
+This model was trained with SFT.
+
+### Framework versions
+
+- TRL: 0.28.0.dev0
+- Transformers: 5.3.0
+- Pytorch: 2.8.0+cu128
+- Datasets: 2.21.0
+- Tokenizers: 0.22.2
+
+## Citations
+
+
+
+Cite TRL as:
+    
+```bibtex
+@software{vonwerra2020trl,
+  title   = {{TRL: Transformers Reinforcement Learning}},
+  author  = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
+  license = {Apache-2.0},
+  url     = {https://github.com/huggingface/trl},
+  year    = {2020}
+}
+```
\ No newline at end of file
diff --git a/chat_template.jinja b/chat_template.jinja
new file mode 100644
index 0000000..28028c0
--- /dev/null
+++ b/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/config.json b/config.json
new file mode 100644
index 0000000..722b1cf
--- /dev/null
+++ b/config.json
@@ -0,0 +1,62 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 131072,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000.0,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.3.0",
+  "use_cache": false,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151680
+}
diff --git a/generation_config.json b/generation_config.json
new file mode 100644
index 0000000..791a05b
--- /dev/null
+++ b/generation_config.json
@@ -0,0 +1,10 @@
+{
+  "do_sample": false,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "max_new_tokens": 2048,
+  "pad_token_id": 151643,
+  "transformers_version": "5.3.0"
+}
diff --git a/model.safetensors b/model.safetensors
new file mode 100644
index 0000000..30e5d8f
--- /dev/null
+++ b/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1c621453202ebd3186134a8697bfa812a2e78ac6f6b49177723b55784673c21
+size 3086680712
diff --git a/rng_state.pth b/rng_state.pth
new file mode 100644
index 0000000..c1cd5ab
--- /dev/null
+++ b/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c68c179797431af2ccf164d3e853ff12fcc4859a5e76600bc4d387eebd54aa2e
+size 14645
diff --git a/scheduler.pt b/scheduler.pt
new file mode 100644
index 0000000..32fca79
--- /dev/null
+++ b/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d66ebfaf80864e59b310ea0321988c04ca90eb65a3d5e188c93a9e636ef920d
+size 1465
diff --git a/tokenizer.json b/tokenizer.json
new file mode 100644
index 0000000..aae1904
--- /dev/null
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78e38431a079a9fd59f593c0496daaf1a945ced9e039fc3223fb02fd36d42e52
+size 11424732
diff --git a/tokenizer_config.json b/tokenizer_config.json
new file mode 100644
index 0000000..7d75d3b
--- /dev/null
+++ b/tokenizer_config.json
@@ -0,0 +1,29 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/trainer_state.json b/trainer_state.json
new file mode 100644
index 0000000..b076ccf
--- /dev/null
+++ b/trainer_state.json
@@ -0,0 +1,1721 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 50,
+  "global_step": 1392,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.6767837120220066,
+      "epoch": 0.021570610043815303,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 4.9994842949609844e-06,
+      "loss": 1.4010092735290527,
+      "mean_token_accuracy": 0.6483692195266485,
+      "num_tokens": 745433.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.6610254392027854,
+      "epoch": 0.043141220087630605,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 4.99770188402723e-06,
+      "loss": 1.3603099822998046,
+      "mean_token_accuracy": 0.6535671677440404,
+      "num_tokens": 1484450.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.638558079302311,
+      "epoch": 0.06471183013144591,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 4.994647308096509e-06,
+      "loss": 1.3407745361328125,
+      "mean_token_accuracy": 0.6570096373558044,
+      "num_tokens": 2243238.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.6027556115761399,
+      "epoch": 0.08628244017526121,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 4.99032212296882e-06,
+      "loss": 1.3085095405578613,
+      "mean_token_accuracy": 0.6618063950911164,
+      "num_tokens": 3001195.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.6135555064305662,
+      "epoch": 0.10785305021907651,
+      "grad_norm": 0.041015625,
+      "learning_rate": 4.984728531608879e-06,
+      "loss": 1.3027904510498047,
+      "mean_token_accuracy": 0.6617284880951046,
+      "num_tokens": 3752781.0,
+      "step": 50
+    },
+    {
+      "epoch": 0.10785305021907651,
+      "eval_entropy": 1.499487464244549,
+      "eval_loss": 1.2577375173568726,
+      "eval_mean_token_accuracy": 0.6712858126713679,
+      "eval_num_tokens": 3752781.0,
+      "eval_runtime": 4.0938,
+      "eval_samples_per_second": 24.427,
+      "eval_steps_per_second": 3.176,
+      "step": 50
+    },
+    {
+      "entropy": 1.5552376886829733,
+      "epoch": 0.12942366026289182,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 4.977869383024079e-06,
+      "loss": 1.2729825019836425,
+      "mean_token_accuracy": 0.6678385319188237,
+      "num_tokens": 4495508.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.6034994365647435,
+      "epoch": 0.15099427030670712,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 4.969748170813396e-06,
+      "loss": 1.3100390434265137,
+      "mean_token_accuracy": 0.6599523894488811,
+      "num_tokens": 5247625.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.5726647214964031,
+      "epoch": 0.17256488035052242,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 4.9603690313879716e-06,
+      "loss": 1.276181125640869,
+      "mean_token_accuracy": 0.6651923535391688,
+      "num_tokens": 5984349.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.5922515178099275,
+      "epoch": 0.19413549039433772,
+      "grad_norm": 0.041015625,
+      "learning_rate": 4.949736741864305e-06,
+      "loss": 1.3018789291381836,
+      "mean_token_accuracy": 0.6616464355960489,
+      "num_tokens": 6741339.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.556330101378262,
+      "epoch": 0.21570610043815303,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 4.9378567176311024e-06,
+      "loss": 1.2867472648620606,
+      "mean_token_accuracy": 0.6638703044503927,
+      "num_tokens": 7499467.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.21570610043815303,
+      "eval_entropy": 1.4564316364435048,
+      "eval_loss": 1.2318497896194458,
+      "eval_mean_token_accuracy": 0.6765855183968177,
+      "eval_num_tokens": 7499467.0,
+      "eval_runtime": 3.6785,
+      "eval_samples_per_second": 27.185,
+      "eval_steps_per_second": 3.534,
+      "step": 100
+    },
+    {
+      "entropy": 1.590408154949546,
+      "epoch": 0.23727671048196833,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 4.924735009591028e-06,
+      "loss": 1.2886395454406738,
+      "mean_token_accuracy": 0.6622924569994211,
+      "num_tokens": 8232251.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.5395626040175556,
+      "epoch": 0.25884732052578363,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 4.910378301078783e-06,
+      "loss": 1.2639349937438964,
+      "mean_token_accuracy": 0.6681265501305461,
+      "num_tokens": 8975352.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.5464642237871886,
+      "epoch": 0.2804179305695989,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 4.894793904457047e-06,
+      "loss": 1.2817666053771972,
+      "mean_token_accuracy": 0.6655796402134001,
+      "num_tokens": 9728467.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.5411764241755008,
+      "epoch": 0.30198854061341424,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 4.877989757392042e-06,
+      "loss": 1.2660977363586425,
+      "mean_token_accuracy": 0.6667622409760952,
+      "num_tokens": 10485295.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.5560635298490524,
+      "epoch": 0.3235591506572295,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 4.859974418810612e-06,
+      "loss": 1.275891399383545,
+      "mean_token_accuracy": 0.6646549022756517,
+      "num_tokens": 11241265.0,
+      "step": 150
+    },
+    {
+      "epoch": 0.3235591506572295,
+      "eval_entropy": 1.43433350783128,
+      "eval_loss": 1.2188570499420166,
+      "eval_mean_token_accuracy": 0.6790601794536297,
+      "eval_num_tokens": 11241265.0,
+      "eval_runtime": 3.6641,
+      "eval_samples_per_second": 27.292,
+      "eval_steps_per_second": 3.548,
+      "step": 150
+    },
+    {
+      "entropy": 1.5227545902132988,
+      "epoch": 0.34512976070104484,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 4.840757064540862e-06,
+      "loss": 1.265657615661621,
+      "mean_token_accuracy": 0.6681289479136467,
+      "num_tokens": 11979175.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.5231037836521863,
+      "epoch": 0.3667003707448601,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 4.820347482638604e-06,
+      "loss": 1.2708560943603515,
+      "mean_token_accuracy": 0.6657222095876932,
+      "num_tokens": 12746440.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.5557100439444185,
+      "epoch": 0.38827098078867545,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 4.798756068401956e-06,
+      "loss": 1.2672907829284668,
+      "mean_token_accuracy": 0.6689961737021803,
+      "num_tokens": 13483465.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.5149108780547977,
+      "epoch": 0.4098415908324907,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 4.775993819076669e-06,
+      "loss": 1.2582406044006347,
+      "mean_token_accuracy": 0.6681622717529535,
+      "num_tokens": 14242169.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.5303620437160135,
+      "epoch": 0.43141220087630605,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 4.752072328254848e-06,
+      "loss": 1.2645947456359863,
+      "mean_token_accuracy": 0.6671571815386415,
+      "num_tokens": 14984546.0,
+      "step": 200
+    },
+    {
+      "epoch": 0.43141220087630605,
+      "eval_entropy": 1.4211730040036714,
+      "eval_loss": 1.2098078727722168,
+      "eval_mean_token_accuracy": 0.6801219215759864,
+      "eval_num_tokens": 14984546.0,
+      "eval_runtime": 3.6827,
+      "eval_samples_per_second": 27.154,
+      "eval_steps_per_second": 3.53,
+      "step": 200
+    },
+    {
+      "entropy": 1.5229917043820023,
+      "epoch": 0.4529828109201213,
+      "grad_norm": 0.03515625,
+      "learning_rate": 4.72700377996994e-06,
+      "loss": 1.252067756652832,
+      "mean_token_accuracy": 0.6694514124654234,
+      "num_tokens": 15735842.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.528000575862825,
+      "epoch": 0.47455342096393666,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 4.700800942490991e-06,
+      "loss": 1.2488290786743164,
+      "mean_token_accuracy": 0.6729495018720627,
+      "num_tokens": 16464563.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.5040436958894134,
+      "epoch": 0.49612403100775193,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 4.673477161819331e-06,
+      "loss": 1.2432649612426758,
+      "mean_token_accuracy": 0.6707049548625946,
+      "num_tokens": 17210962.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.5237091314047575,
+      "epoch": 0.5176946410515673,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 4.645046354890994e-06,
+      "loss": 1.2539048194885254,
+      "mean_token_accuracy": 0.66995022688061,
+      "num_tokens": 17957358.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.5211739281192422,
+      "epoch": 0.5392652510953826,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 4.615523002488352e-06,
+      "loss": 1.2501190185546875,
+      "mean_token_accuracy": 0.6701988697052002,
+      "num_tokens": 18699427.0,
+      "step": 250
+    },
+    {
+      "epoch": 0.5392652510953826,
+      "eval_entropy": 1.4076783198576708,
+      "eval_loss": 1.203089714050293,
+      "eval_mean_token_accuracy": 0.6810761552590591,
+      "eval_num_tokens": 18699427.0,
+      "eval_runtime": 3.6577,
+      "eval_samples_per_second": 27.339,
+      "eval_steps_per_second": 3.554,
+      "step": 250
+    },
+    {
+      "entropy": 1.5334273751825094,
+      "epoch": 0.5608358611391978,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 4.584922141864553e-06,
+      "loss": 1.2668370246887206,
+      "mean_token_accuracy": 0.6666171688586473,
+      "num_tokens": 19455215.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.521253040060401,
+      "epoch": 0.5824064711830131,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 4.5532593590845345e-06,
+      "loss": 1.2646316528320312,
+      "mean_token_accuracy": 0.6671235594898463,
+      "num_tokens": 20186200.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.5162998696789145,
+      "epoch": 0.6039770812268285,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 4.520550781086511e-06,
+      "loss": 1.2609230041503907,
+      "mean_token_accuracy": 0.6699131606146693,
+      "num_tokens": 20928161.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.5239347798749805,
+      "epoch": 0.6255476912706438,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 4.48681306746797e-06,
+      "loss": 1.2647696495056153,
+      "mean_token_accuracy": 0.6664841236546636,
+      "num_tokens": 21677030.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.5348559882491828,
+      "epoch": 0.647118301314459,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 4.452063402000369e-06,
+      "loss": 1.2626654624938964,
+      "mean_token_accuracy": 0.6681625910103322,
+      "num_tokens": 22441632.0,
+      "step": 300
+    },
+    {
+      "epoch": 0.647118301314459,
+      "eval_entropy": 1.400939602118272,
+      "eval_loss": 1.1979954242706299,
+      "eval_mean_token_accuracy": 0.6817522003100469,
+      "eval_num_tokens": 22441632.0,
+      "eval_runtime": 3.6816,
+      "eval_samples_per_second": 27.162,
+      "eval_steps_per_second": 3.531,
+      "step": 300
+    },
+    {
+      "entropy": 1.4912159774452447,
+      "epoch": 0.6686889113582744,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 4.416319483876849e-06,
+      "loss": 1.235336971282959,
+      "mean_token_accuracy": 0.6746378557756543,
+      "num_tokens": 23183185.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.5201040310785174,
+      "epoch": 0.6902595214020897,
+      "grad_norm": 0.033203125,
+      "learning_rate": 4.379599518697444e-06,
+      "loss": 1.2683670043945312,
+      "mean_token_accuracy": 0.6668972001411021,
+      "num_tokens": 23926617.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.5076332142576576,
+      "epoch": 0.7118301314459049,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 4.341922209196327e-06,
+      "loss": 1.261031150817871,
+      "mean_token_accuracy": 0.6689906693994999,
+      "num_tokens": 24694574.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.48094400819391,
+      "epoch": 0.7334007414897202,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 4.303306745715885e-06,
+      "loss": 1.2272047042846679,
+      "mean_token_accuracy": 0.675537558272481,
+      "num_tokens": 25431819.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.4832843391224741,
+      "epoch": 0.7549713515335356,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 4.263772796432409e-06,
+      "loss": 1.229203987121582,
+      "mean_token_accuracy": 0.6737644655629993,
+      "num_tokens": 26168439.0,
+      "step": 350
+    },
+    {
+      "epoch": 0.7549713515335356,
+      "eval_entropy": 1.3908999791512122,
+      "eval_loss": 1.1945923566818237,
+      "eval_mean_token_accuracy": 0.6822104912537795,
+      "eval_num_tokens": 26168439.0,
+      "eval_runtime": 3.6517,
+      "eval_samples_per_second": 27.384,
+      "eval_steps_per_second": 3.56,
+      "step": 350
+    },
+    {
+      "entropy": 1.4919319538399578,
+      "epoch": 0.7765419615773509,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 4.223340497338425e-06,
+      "loss": 1.254155445098877,
+      "mean_token_accuracy": 0.669323449395597,
+      "num_tokens": 26912893.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.4747931480407714,
+      "epoch": 0.7981125716211661,
+      "grad_norm": 0.03515625,
+      "learning_rate": 4.182030441986744e-06,
+      "loss": 1.2301451683044433,
+      "mean_token_accuracy": 0.6740846395492553,
+      "num_tokens": 27654040.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.5113066220656037,
+      "epoch": 0.8196831816649814,
+      "grad_norm": 0.03515625,
+      "learning_rate": 4.139863671001462e-06,
+      "loss": 1.2547582626342773,
+      "mean_token_accuracy": 0.668442091718316,
+      "num_tokens": 28415074.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.4829508623108267,
+      "epoch": 0.8412537917087968,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 4.096861661361244e-06,
+      "loss": 1.2368510246276856,
+      "mean_token_accuracy": 0.6739088529720902,
+      "num_tokens": 29164236.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.4742595052346588,
+      "epoch": 0.8628244017526121,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 4.053046315460375e-06,
+      "loss": 1.221826934814453,
+      "mean_token_accuracy": 0.6764747254550457,
+      "num_tokens": 29917850.0,
+      "step": 400
+    },
+    {
+      "epoch": 0.8628244017526121,
+      "eval_entropy": 1.3889273129976714,
+      "eval_loss": 1.1914775371551514,
+      "eval_mean_token_accuracy": 0.6828073492416968,
+      "eval_num_tokens": 29917850.0,
+      "eval_runtime": 3.6791,
+      "eval_samples_per_second": 27.18,
+      "eval_steps_per_second": 3.533,
+      "step": 400
+    },
+    {
+      "entropy": 1.5197141887620091,
+      "epoch": 0.8843950117964273,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 4.008439949953111e-06,
+      "loss": 1.2676182746887208,
+      "mean_token_accuracy": 0.6672423681244254,
+      "num_tokens": 30652050.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.4985390733927488,
+      "epoch": 0.9059656218402427,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 3.96306528438705e-06,
+      "loss": 1.2611710548400878,
+      "mean_token_accuracy": 0.6676615543663502,
+      "num_tokens": 31397271.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.4782635426148771,
+      "epoch": 0.927536231884058,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 3.9169454296312895e-06,
+      "loss": 1.2356756210327149,
+      "mean_token_accuracy": 0.6728357087820769,
+      "num_tokens": 32137148.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.5283029813319444,
+      "epoch": 0.9491068419278733,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 3.870103876105266e-06,
+      "loss": 1.261199378967285,
+      "mean_token_accuracy": 0.6644394308328628,
+      "num_tokens": 32883234.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.4918072402477265,
+      "epoch": 0.9706774519716885,
+      "grad_norm": 0.03173828125,
+      "learning_rate": 3.822564481814285e-06,
+      "loss": 1.2366758346557618,
+      "mean_token_accuracy": 0.6730084288865328,
+      "num_tokens": 33621739.0,
+      "step": 450
+    },
+    {
+      "epoch": 0.9706774519716885,
+      "eval_entropy": 1.3830306346599872,
+      "eval_loss": 1.189202904701233,
+      "eval_mean_token_accuracy": 0.6829607074077313,
+      "eval_num_tokens": 33621739.0,
+      "eval_runtime": 3.6568,
+      "eval_samples_per_second": 27.346,
+      "eval_steps_per_second": 3.555,
+      "step": 450
+    },
+    {
+      "entropy": 1.4637328485026955,
+      "epoch": 0.9922480620155039,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 3.7743514601978194e-06,
+      "loss": 1.2146634101867675,
+      "mean_token_accuracy": 0.6782162645831704,
+      "num_tokens": 34370732.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.4961490114659362,
+      "epoch": 1.0129423660262893,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 3.7254893677967817e-06,
+      "loss": 1.2408208847045898,
+      "mean_token_accuracy": 0.6726808932394469,
+      "num_tokens": 35092019.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.4762530026957392,
+      "epoch": 1.0345129760701044,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 3.6760030917460327e-06,
+      "loss": 1.2347814559936523,
+      "mean_token_accuracy": 0.6722965622320771,
+      "num_tokens": 35842937.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.5049012653529643,
+      "epoch": 1.0560835861139197,
+      "grad_norm": 0.0390625,
+      "learning_rate": 3.6259178370985143e-06,
+      "loss": 1.2423776626586913,
+      "mean_token_accuracy": 0.6726679367944598,
+      "num_tokens": 36586521.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.5029818255454301,
+      "epoch": 1.077654196157735,
+      "grad_norm": 0.0390625,
+      "learning_rate": 3.5752591139874555e-06,
+      "loss": 1.2592316627502442,
+      "mean_token_accuracy": 0.6691493924707175,
+      "num_tokens": 37334170.0,
+      "step": 500
+    },
+    {
+      "epoch": 1.077654196157735,
+      "eval_entropy": 1.380214232664842,
+      "eval_loss": 1.187309980392456,
+      "eval_mean_token_accuracy": 0.6840003041120676,
+      "eval_num_tokens": 37334170.0,
+      "eval_runtime": 3.6676,
+      "eval_samples_per_second": 27.265,
+      "eval_steps_per_second": 3.545,
+      "step": 500
+    },
+    {
+      "entropy": 1.5072425819933415,
+      "epoch": 1.0992248062015504,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 3.5240527246331875e-06,
+      "loss": 1.2472517013549804,
+      "mean_token_accuracy": 0.6706916518509388,
+      "num_tokens": 38077388.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.4855062494054436,
+      "epoch": 1.1207954162453657,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 3.47232475020119e-06,
+      "loss": 1.2272147178649901,
+      "mean_token_accuracy": 0.674690573848784,
+      "num_tokens": 38850755.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.5008873930200934,
+      "epoch": 1.142366026289181,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 3.4201015375180636e-06,
+      "loss": 1.2496809005737304,
+      "mean_token_accuracy": 0.6694020047783852,
+      "num_tokens": 39611349.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.4931515300646425,
+      "epoch": 1.1639366363329964,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 3.3674096856521866e-06,
+      "loss": 1.2493350028991699,
+      "mean_token_accuracy": 0.6695653740316629,
+      "num_tokens": 40345796.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.4781431961804627,
+      "epoch": 1.1855072463768117,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 3.3142760323659035e-06,
+      "loss": 1.2255599975585938,
+      "mean_token_accuracy": 0.6748269772157073,
+      "num_tokens": 41092178.0,
+      "step": 550
+    },
+    {
+      "epoch": 1.1855072463768117,
+      "eval_entropy": 1.374479284653297,
+      "eval_loss": 1.185532808303833,
+      "eval_mean_token_accuracy": 0.683243086704841,
+      "eval_num_tokens": 41092178.0,
+      "eval_runtime": 3.6823,
+      "eval_samples_per_second": 27.157,
+      "eval_steps_per_second": 3.53,
+      "step": 550
+    },
+    {
+      "entropy": 1.4843646846711636,
+      "epoch": 1.2070778564206268,
+      "grad_norm": 0.03515625,
+      "learning_rate": 3.2607276404461325e-06,
+      "loss": 1.2443278312683106,
+      "mean_token_accuracy": 0.6696186825633049,
+      "num_tokens": 41845174.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.489796095713973,
+      "epoch": 1.2286484664644421,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 3.206791783920361e-06,
+      "loss": 1.2421024322509766,
+      "mean_token_accuracy": 0.670965776592493,
+      "num_tokens": 42586335.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.4751165067777037,
+      "epoch": 1.2502190765082575,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 3.152495934165055e-06,
+      "loss": 1.2281587600708008,
+      "mean_token_accuracy": 0.6753974489867687,
+      "num_tokens": 43306996.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.5005261640995742,
+      "epoch": 1.2717896865520728,
+      "grad_norm": 0.033203125,
+      "learning_rate": 3.0978677459135443e-06,
+      "loss": 1.2365185737609863,
+      "mean_token_accuracy": 0.6739299269393086,
+      "num_tokens": 44076004.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.4971575114876032,
+      "epoch": 1.2933602965958881,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 3.0429350431705203e-06,
+      "loss": 1.2424898147583008,
+      "mean_token_accuracy": 0.6698520433157682,
+      "num_tokens": 44826409.0,
+      "step": 600
+    },
+    {
+      "epoch": 1.2933602965958881,
+      "eval_entropy": 1.3752018488370454,
+      "eval_loss": 1.1843688488006592,
+      "eval_mean_token_accuracy": 0.6843544657413776,
+      "eval_num_tokens": 44826409.0,
+      "eval_runtime": 3.6729,
+      "eval_samples_per_second": 27.226,
+      "eval_steps_per_second": 3.539,
+      "step": 600
+    },
+    {
+      "entropy": 1.4854857932776213,
+      "epoch": 1.3149309066397035,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.9877258050403214e-06,
+      "loss": 1.2554213523864746,
+      "mean_token_accuracy": 0.6699382528662682,
+      "num_tokens": 45581714.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.501467790082097,
+      "epoch": 1.3365015166835188,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 2.932268151476216e-06,
+      "loss": 1.2491737365722657,
+      "mean_token_accuracy": 0.6689437933266162,
+      "num_tokens": 46333632.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.4766393253579735,
+      "epoch": 1.358072126727334,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.8765903289579504e-06,
+      "loss": 1.2321022987365722,
+      "mean_token_accuracy": 0.6739912224933505,
+      "num_tokens": 47079975.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.5021409541368484,
+      "epoch": 1.3796427367711492,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 2.8207206961048496e-06,
+      "loss": 1.2492496490478515,
+      "mean_token_accuracy": 0.6727896144613623,
+      "num_tokens": 47840346.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.478764550201595,
+      "epoch": 1.4012133468149646,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 2.7646877092318005e-06,
+      "loss": 1.2151342391967774,
+      "mean_token_accuracy": 0.6766992717981338,
+      "num_tokens": 48578138.0,
+      "step": 650
+    },
+    {
+      "epoch": 1.4012133468149646,
+      "eval_entropy": 1.373001291201665,
+      "eval_loss": 1.1834796667099,
+      "eval_mean_token_accuracy": 0.6841690998810989,
+      "eval_num_tokens": 48578138.0,
+      "eval_runtime": 3.6584,
+      "eval_samples_per_second": 27.334,
+      "eval_steps_per_second": 3.553,
+      "step": 650
+    },
+    {
+      "entropy": 1.4845669204369187,
+      "epoch": 1.42278395685878,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.7085199078554816e-06,
+      "loss": 1.229975986480713,
+      "mean_token_accuracy": 0.6744854656979442,
+      "num_tokens": 49327776.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.4945817491039635,
+      "epoch": 1.4443545669025952,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.652245900158208e-06,
+      "loss": 1.2332301139831543,
+      "mean_token_accuracy": 0.6740596545860171,
+      "num_tokens": 50088883.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.4762866148725151,
+      "epoch": 1.4659251769464106,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.595894348416807e-06,
+      "loss": 1.2285600662231446,
+      "mean_token_accuracy": 0.6744002403691411,
+      "num_tokens": 50844287.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.4770476857200265,
+      "epoch": 1.4874957869902259,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 2.539493954403939e-06,
+      "loss": 1.2145397186279296,
+      "mean_token_accuracy": 0.6767452768981457,
+      "num_tokens": 51595679.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.4556858593598008,
+      "epoch": 1.509066397034041,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 2.483073444769304e-06,
+      "loss": 1.2092321395874024,
+      "mean_token_accuracy": 0.6777302039787173,
+      "num_tokens": 52334062.0,
+      "step": 700
+    },
+    {
+      "epoch": 1.509066397034041,
+      "eval_entropy": 1.3710510272246141,
+      "eval_loss": 1.1827597618103027,
+      "eval_mean_token_accuracy": 0.6842976808547974,
+      "eval_num_tokens": 52334062.0,
+      "eval_runtime": 3.6872,
+      "eval_samples_per_second": 27.121,
+      "eval_steps_per_second": 3.526,
+      "step": 700
+    },
+    {
+      "entropy": 1.499366302229464,
+      "epoch": 1.5306370070778565,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 2.426661556408178e-06,
+      "loss": 1.2455547332763672,
+      "mean_token_accuracy": 0.670172212831676,
+      "num_tokens": 53075155.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.4780711084604263,
+      "epoch": 1.5522076171216717,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 2.370287021824725e-06,
+      "loss": 1.2432132720947267,
+      "mean_token_accuracy": 0.6701041288673878,
+      "num_tokens": 53824562.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.477502679079771,
+      "epoch": 1.573778227165487,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.3139785544975528e-06,
+      "loss": 1.2311717987060546,
+      "mean_token_accuracy": 0.6731598127633334,
+      "num_tokens": 54581099.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.469127612374723,
+      "epoch": 1.5953488372093023,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.2577648342549577e-06,
+      "loss": 1.2264204978942872,
+      "mean_token_accuracy": 0.6752710843458771,
+      "num_tokens": 55327112.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.4831633877009154,
+      "epoch": 1.6169194472531176,
+      "grad_norm": 0.033203125,
+      "learning_rate": 2.201674492667303e-06,
+      "loss": 1.2290393829345703,
+      "mean_token_accuracy": 0.6734487740322948,
+      "num_tokens": 56077303.0,
+      "step": 750
+    },
+    {
+      "epoch": 1.6169194472531176,
+      "eval_entropy": 1.3741891384124756,
+      "eval_loss": 1.1823630332946777,
+      "eval_mean_token_accuracy": 0.683963784804711,
+      "eval_num_tokens": 56077303.0,
+      "eval_runtime": 3.6887,
+      "eval_samples_per_second": 27.11,
+      "eval_steps_per_second": 3.524,
+      "step": 750
+    },
+    {
+      "entropy": 1.4855883978307247,
+      "epoch": 1.638490057296933,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.145736098463985e-06,
+      "loss": 1.2290058135986328,
+      "mean_token_accuracy": 0.6753679029643536,
+      "num_tokens": 56812770.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.4715428488329052,
+      "epoch": 1.660060667340748,
+      "grad_norm": 0.03515625,
+      "learning_rate": 2.0899781429823994e-06,
+      "loss": 1.219193172454834,
+      "mean_token_accuracy": 0.6744225597009063,
+      "num_tokens": 57559964.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.50134886931628,
+      "epoch": 1.6816312773845636,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.03442902565633e-06,
+      "loss": 1.241888427734375,
+      "mean_token_accuracy": 0.6697187574580312,
+      "num_tokens": 58308737.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.4558339582756161,
+      "epoch": 1.7032018874283787,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 1.979117039551144e-06,
+      "loss": 1.2013543128967286,
+      "mean_token_accuracy": 0.6808343829587102,
+      "num_tokens": 59052112.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.4817260129377245,
+      "epoch": 1.7247724974721943,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.924070356953168e-06,
+      "loss": 1.2334010124206543,
+      "mean_token_accuracy": 0.672742473334074,
+      "num_tokens": 59790196.0,
+      "step": 800
+    },
+    {
+      "epoch": 1.7247724974721943,
+      "eval_entropy": 1.37288563985091,
+      "eval_loss": 1.1820176839828491,
+      "eval_mean_token_accuracy": 0.684046956209036,
+      "eval_num_tokens": 59790196.0,
+      "eval_runtime": 3.6779,
+      "eval_samples_per_second": 27.19,
+      "eval_steps_per_second": 3.535,
+      "step": 800
+    },
+    {
+      "entropy": 1.4934301238507033,
+      "epoch": 1.7463431075160094,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.8693170150205791e-06,
+      "loss": 1.2414652824401855,
+      "mean_token_accuracy": 0.672171187773347,
+      "num_tokens": 60528075.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.4653095223009587,
+      "epoch": 1.7679137175598247,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.8148849015031195e-06,
+      "loss": 1.2253877639770507,
+      "mean_token_accuracy": 0.6749245407059788,
+      "num_tokens": 61270945.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.470803551748395,
+      "epoch": 1.78948432760364,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 1.7608017405379133e-06,
+      "loss": 1.2263381004333496,
+      "mean_token_accuracy": 0.6744035631418228,
+      "num_tokens": 62016952.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.464294382557273,
+      "epoch": 1.8110549376474552,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.707095078528615e-06,
+      "loss": 1.2164762496948243,
+      "mean_token_accuracy": 0.6769138135015964,
+      "num_tokens": 62749488.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.471432145871222,
+      "epoch": 1.8326255476912707,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.6537922701150826e-06,
+      "loss": 1.217400360107422,
+      "mean_token_accuracy": 0.677493049018085,
+      "num_tokens": 63488076.0,
+      "step": 850
+    },
+    {
+      "epoch": 1.8326255476912707,
+      "eval_entropy": 1.3718278774848351,
+      "eval_loss": 1.1815452575683594,
+      "eval_mean_token_accuracy": 0.6844739180344802,
+      "eval_num_tokens": 63488076.0,
+      "eval_runtime": 3.6567,
+      "eval_samples_per_second": 27.347,
+      "eval_steps_per_second": 3.555,
+      "step": 850
+    },
+    {
+      "entropy": 1.4581330316141248,
+      "epoch": 1.8541961577350858,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 1.6009204642407269e-06,
+      "loss": 1.2144439697265625,
+      "mean_token_accuracy": 0.6787769604474306,
+      "num_tokens": 64230870.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.4812420466914773,
+      "epoch": 1.8757667677789014,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.548506590324622e-06,
+      "loss": 1.2369457244873048,
+      "mean_token_accuracy": 0.6696851562708617,
+      "num_tokens": 64965408.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.461279028095305,
+      "epoch": 1.8973373778227165,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.4965773445454348e-06,
+      "loss": 1.2258675575256348,
+      "mean_token_accuracy": 0.6737756703048945,
+      "num_tokens": 65706011.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.478666521050036,
+      "epoch": 1.9189079878665318,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 1.4451591762441461e-06,
+      "loss": 1.234369945526123,
+      "mean_token_accuracy": 0.6738810567185283,
+      "num_tokens": 66452496.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.4450379524379968,
+      "epoch": 1.9404785979103472,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.3942782744524974e-06,
+      "loss": 1.1851463317871094,
+      "mean_token_accuracy": 0.6811737598851323,
+      "num_tokens": 67202492.0,
+      "step": 900
+    },
+    {
+      "epoch": 1.9404785979103472,
+      "eval_entropy": 1.372688889503479,
+      "eval_loss": 1.1814879179000854,
+      "eval_mean_token_accuracy": 0.6845757319377019,
+      "eval_num_tokens": 67202492.0,
+      "eval_runtime": 3.6701,
+      "eval_samples_per_second": 27.247,
+      "eval_steps_per_second": 3.542,
+      "step": 900
+    },
+    {
+      "entropy": 1.4832466956228019,
+      "epoch": 1.9620492079541625,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.343960554554019e-06,
+      "loss": 1.2365760803222656,
+      "mean_token_accuracy": 0.6740058463066816,
+      "num_tokens": 67959083.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.4821348998695612,
+      "epoch": 1.9836198179979778,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 1.2942316450844455e-06,
+      "loss": 1.2364745140075684,
+      "mean_token_accuracy": 0.6715554362162948,
+      "num_tokens": 68704193.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.4634444566037057,
+      "epoch": 2.004314122008763,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 1.2451168746782154e-06,
+      "loss": 1.2172240257263183,
+      "mean_token_accuracy": 0.6751755258547755,
+      "num_tokens": 69432162.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.4495740208774806,
+      "epoch": 2.0258847320525786,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.1966412591677432e-06,
+      "loss": 1.2037148475646973,
+      "mean_token_accuracy": 0.6786413632333279,
+      "num_tokens": 70174260.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.498362773656845,
+      "epoch": 2.0474553420963937,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.1488294888419897e-06,
+      "loss": 1.248203182220459,
+      "mean_token_accuracy": 0.6689321782439948,
+      "num_tokens": 70913892.0,
+      "step": 950
+    },
+    {
+      "epoch": 2.0474553420963937,
+      "eval_entropy": 1.3701610014988825,
+      "eval_loss": 1.18141770362854,
+      "eval_mean_token_accuracy": 0.6844954398962168,
+      "eval_num_tokens": 70913892.0,
+      "eval_runtime": 3.6782,
+      "eval_samples_per_second": 27.187,
+      "eval_steps_per_second": 3.534,
+      "step": 950
+    },
+    {
+      "entropy": 1.4788924545049666,
+      "epoch": 2.069025952140209,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.1017059158708595e-06,
+      "loss": 1.2327592849731446,
+      "mean_token_accuracy": 0.6722837900742888,
+      "num_tokens": 71658336.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.4883825523778795,
+      "epoch": 2.0905965621840243,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 1.0552945419018066e-06,
+      "loss": 1.2401643753051759,
+      "mean_token_accuracy": 0.6708007922396064,
+      "num_tokens": 72411055.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.4376504089683295,
+      "epoch": 2.1121671722278395,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.0096190058349753e-06,
+      "loss": 1.2030488014221192,
+      "mean_token_accuracy": 0.6795810239389539,
+      "num_tokens": 73159842.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.4649095240980388,
+      "epoch": 2.133737782271655,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 9.647025717830967e-07,
+      "loss": 1.211749267578125,
+      "mean_token_accuracy": 0.6779318297281861,
+      "num_tokens": 73903738.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.4727241741493344,
+      "epoch": 2.15530839231547,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 9.205681172222855e-07,
+      "loss": 1.2229228973388673,
+      "mean_token_accuracy": 0.6762760354205966,
+      "num_tokens": 74631034.0,
+      "step": 1000
+    },
+    {
+      "epoch": 2.15530839231547,
+      "eval_entropy": 1.3697441999728863,
+      "eval_loss": 1.181374430656433,
+      "eval_mean_token_accuracy": 0.6845315419710599,
+      "eval_num_tokens": 74631034.0,
+      "eval_runtime": 3.6618,
+      "eval_samples_per_second": 27.309,
+      "eval_steps_per_second": 3.55,
+      "step": 1000
+    },
+    {
+      "entropy": 1.4828446479514241,
+      "epoch": 2.1768790023592857,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 8.772381213397527e-07,
+      "loss": 1.2276836395263673,
+      "mean_token_accuracy": 0.6742623111233115,
+      "num_tokens": 75381772.0,
+      "step": 1010
+    },
+    {
+      "entropy": 1.470299855619669,
+      "epoch": 2.1984496124031008,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 8.347346535843931e-07,
+      "loss": 1.2108376502990723,
+      "mean_token_accuracy": 0.6764363253489136,
+      "num_tokens": 76116800.0,
+      "step": 1020
+    },
+    {
+      "entropy": 1.4736272709444165,
+      "epoch": 2.220020222446916,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 7.930793624260542e-07,
+      "loss": 1.2280489921569824,
+      "mean_token_accuracy": 0.6724860800430179,
+      "num_tokens": 76863722.0,
+      "step": 1030
+    },
+    {
+      "entropy": 1.5003264799714089,
+      "epoch": 2.2415908324907314,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 7.522934643292357e-07,
+      "loss": 1.2437036514282227,
+      "mean_token_accuracy": 0.6705104760825634,
+      "num_tokens": 77614440.0,
+      "step": 1040
+    },
+    {
+      "entropy": 1.4879769939929246,
+      "epoch": 2.2631614425345465,
+      "grad_norm": 0.03515625,
+      "learning_rate": 7.123977329468129e-07,
+      "loss": 1.242294979095459,
+      "mean_token_accuracy": 0.6722773084416985,
+      "num_tokens": 78373949.0,
+      "step": 1050
+    },
+    {
+      "epoch": 2.2631614425345465,
+      "eval_entropy": 1.3714183202156653,
+      "eval_loss": 1.1815378665924072,
+      "eval_mean_token_accuracy": 0.6843903064727783,
+      "eval_num_tokens": 78373949.0,
+      "eval_runtime": 3.6971,
+      "eval_samples_per_second": 27.048,
+      "eval_steps_per_second": 3.516,
+      "step": 1050
+    },
+    {
+      "entropy": 1.4705588215962053,
+      "epoch": 2.284732052578362,
+      "grad_norm": 0.033203125,
+      "learning_rate": 6.734124885393111e-07,
+      "loss": 1.2206010818481445,
+      "mean_token_accuracy": 0.6755671601742506,
+      "num_tokens": 79126819.0,
+      "step": 1060
+    },
+    {
+      "entropy": 1.4693676881492137,
+      "epoch": 2.306302662622177,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 6.353575876250945e-07,
+      "loss": 1.2280621528625488,
+      "mean_token_accuracy": 0.6749383028596639,
+      "num_tokens": 79869254.0,
+      "step": 1070
+    },
+    {
+      "entropy": 1.4845634065568447,
+      "epoch": 2.3278732726659928,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 5.982524128667697e-07,
+      "loss": 1.225168800354004,
+      "mean_token_accuracy": 0.6747270798310637,
+      "num_tokens": 80603220.0,
+      "step": 1080
+    },
+    {
+      "entropy": 1.478745934739709,
+      "epoch": 2.349443882709808,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 5.621158631989263e-07,
+      "loss": 1.2239156723022462,
+      "mean_token_accuracy": 0.675847783498466,
+      "num_tokens": 81368355.0,
+      "step": 1090
+    },
+    {
+      "entropy": 1.47794102281332,
+      "epoch": 2.3710144927536234,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 5.269663442022685e-07,
+      "loss": 1.2166450500488282,
+      "mean_token_accuracy": 0.6764153370633721,
+      "num_tokens": 82121841.0,
+      "step": 1100
+    },
+    {
+      "epoch": 2.3710144927536234,
+      "eval_entropy": 1.3739139116727388,
+      "eval_loss": 1.1813687086105347,
+      "eval_mean_token_accuracy": 0.6841404208770165,
+      "eval_num_tokens": 82121841.0,
+      "eval_runtime": 3.6763,
+      "eval_samples_per_second": 27.202,
+      "eval_steps_per_second": 3.536,
+      "step": 1100
+    },
+    {
+      "entropy": 1.47496219240129,
+      "epoch": 2.3925851027974385,
+      "grad_norm": 0.03173828125,
+      "learning_rate": 4.928217587290182e-07,
+      "loss": 1.218714141845703,
+      "mean_token_accuracy": 0.6762027183547616,
+      "num_tokens": 82875329.0,
+      "step": 1110
+    },
+    {
+      "entropy": 1.47492984700948,
+      "epoch": 2.4141557128412536,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 4.5969949778438575e-07,
+      "loss": 1.2254411697387695,
+      "mean_token_accuracy": 0.6733096215873957,
+      "num_tokens": 83619772.0,
+      "step": 1120
+    },
+    {
+      "entropy": 1.4938979675993322,
+      "epoch": 2.435726322885069,
+      "grad_norm": 0.033203125,
+      "learning_rate": 4.2761643166873006e-07,
+      "loss": 1.2486177444458009,
+      "mean_token_accuracy": 0.670450198277831,
+      "num_tokens": 84378938.0,
+      "step": 1130
+    },
+    {
+      "entropy": 1.4558670258149504,
+      "epoch": 2.4572969329288843,
+      "grad_norm": 0.033203125,
+      "learning_rate": 3.9658890138494335e-07,
+      "loss": 1.2235937118530273,
+      "mean_token_accuracy": 0.6756238445639611,
+      "num_tokens": 85114018.0,
+      "step": 1140
+    },
+    {
+      "entropy": 1.4995972203090786,
+      "epoch": 2.4788675429727,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 3.6663271031541493e-07,
+      "loss": 1.2367911338806152,
+      "mean_token_accuracy": 0.6726962340995669,
+      "num_tokens": 85856320.0,
+      "step": 1150
+    },
+    {
+      "epoch": 2.4788675429727,
+      "eval_entropy": 1.3706045242456288,
+      "eval_loss": 1.1811416149139404,
+      "eval_mean_token_accuracy": 0.6845263747068552,
+      "eval_num_tokens": 85856320.0,
+      "eval_runtime": 3.6647,
+      "eval_samples_per_second": 27.287,
+      "eval_steps_per_second": 3.547,
+      "step": 1150
+    },
+    {
+      "entropy": 1.4837047975510358,
+      "epoch": 2.500438153016515,
+      "grad_norm": 0.03515625,
+      "learning_rate": 3.3776311617283035e-07,
+      "loss": 1.2384218215942382,
+      "mean_token_accuracy": 0.6717864293605089,
+      "num_tokens": 86585500.0,
+      "step": 1160
+    },
+    {
+      "entropy": 1.4787415696308017,
+      "epoch": 2.52200876306033,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 3.0999482322889574e-07,
+      "loss": 1.2136137008666992,
+      "mean_token_accuracy": 0.6769431287422776,
+      "num_tokens": 87347776.0,
+      "step": 1170
+    },
+    {
+      "entropy": 1.5027014980092646,
+      "epoch": 2.5435793731041456,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 2.8334197482495107e-07,
+      "loss": 1.253582763671875,
+      "mean_token_accuracy": 0.667684393003583,
+      "num_tokens": 88090036.0,
+      "step": 1180
+    },
+    {
+      "entropy": 1.488570535928011,
+      "epoch": 2.565149983147961,
+      "grad_norm": 0.03515625,
+      "learning_rate": 2.5781814616827936e-07,
+      "loss": 1.2479755401611328,
+      "mean_token_accuracy": 0.6704706002026797,
+      "num_tokens": 88854033.0,
+      "step": 1190
+    },
+    {
+      "entropy": 1.5026418125256895,
+      "epoch": 2.5867205931917763,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.3343633741779255e-07,
+      "loss": 1.2379155158996582,
+      "mean_token_accuracy": 0.6707732306793333,
+      "num_tokens": 89595956.0,
+      "step": 1200
+    },
+    {
+      "epoch": 2.5867205931917763,
+      "eval_entropy": 1.3714110851287842,
+      "eval_loss": 1.1810975074768066,
+      "eval_mean_token_accuracy": 0.6843558412331802,
+      "eval_num_tokens": 89595956.0,
+      "eval_runtime": 3.6633,
+      "eval_samples_per_second": 27.298,
+      "eval_steps_per_second": 3.549,
+      "step": 1200
+    },
+    {
+      "entropy": 1.4757313583046199,
+      "epoch": 2.6082912032355914,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 2.102089670626037e-07,
+      "loss": 1.22061185836792,
+      "mean_token_accuracy": 0.6742894217371941,
+      "num_tokens": 90338273.0,
+      "step": 1210
+    },
+    {
+      "entropy": 1.4955619590356946,
+      "epoch": 2.629861813279407,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 1.8814786559686689e-07,
+      "loss": 1.2501788139343262,
+      "mean_token_accuracy": 0.66965347584337,
+      "num_tokens": 91082715.0,
+      "step": 1220
+    },
+    {
+      "entropy": 1.4938439471647142,
+      "epoch": 2.651432423323222,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.672642694941015e-07,
+      "loss": 1.2471290588378907,
+      "mean_token_accuracy": 0.6704504562541842,
+      "num_tokens": 91830465.0,
+      "step": 1230
+    },
+    {
+      "entropy": 1.4964844120666385,
+      "epoch": 2.6730030333670376,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 1.4756881548407153e-07,
+      "loss": 1.2559056282043457,
+      "mean_token_accuracy": 0.6671904474496841,
+      "num_tokens": 92587225.0,
+      "step": 1240
+    },
+    {
+      "entropy": 1.4705483008176088,
+      "epoch": 2.6945736434108527,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.290715351351371e-07,
+      "loss": 1.2225461959838868,
+      "mean_token_accuracy": 0.674612145125866,
+      "num_tokens": 93329195.0,
+      "step": 1250
+    },
+    {
+      "epoch": 2.6945736434108527,
+      "eval_entropy": 1.3716207284193773,
+      "eval_loss": 1.1809742450714111,
+      "eval_mean_token_accuracy": 0.684914414699261,
+      "eval_num_tokens": 93329195.0,
+      "eval_runtime": 3.6738,
+      "eval_samples_per_second": 27.219,
+      "eval_steps_per_second": 3.539,
+      "step": 1250
+    },
+    {
+      "entropy": 1.473252959549427,
+      "epoch": 2.716144253454668,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.1178184974483114e-07,
+      "loss": 1.2252808570861817,
+      "mean_token_accuracy": 0.6753275262191891,
+      "num_tokens": 94073433.0,
+      "step": 1260
+    },
+    {
+      "entropy": 1.495525497198105,
+      "epoch": 2.7377148634984834,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 9.570856554127206e-08,
+      "loss": 1.2336169242858888,
+      "mean_token_accuracy": 0.6725098475813865,
+      "num_tokens": 94817670.0,
+      "step": 1270
+    },
+    {
+      "entropy": 1.477628206834197,
+      "epoch": 2.7592854735422985,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 8.085986919784933e-08,
+      "loss": 1.2343855857849122,
+      "mean_token_accuracy": 0.6734241990372538,
+      "num_tokens": 95571602.0,
+      "step": 1280
+    },
+    {
+      "entropy": 1.4527918914332987,
+      "epoch": 2.780856083586114,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 6.724332366347258e-08,
+      "loss": 1.219487762451172,
+      "mean_token_accuracy": 0.6743847265839576,
+      "num_tokens": 96321379.0,
+      "step": 1290
+    },
+    {
+      "entropy": 1.4562704721465707,
+      "epoch": 2.802426693629929,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 5.4865864310500725e-08,
+      "loss": 1.210085678100586,
+      "mean_token_accuracy": 0.6762114005163312,
+      "num_tokens": 97057191.0,
+      "step": 1300
+    },
+    {
+      "epoch": 2.802426693629929,
+      "eval_entropy": 1.3706807815111601,
+      "eval_loss": 1.1813178062438965,
+      "eval_mean_token_accuracy": 0.6852421439610995,
+      "eval_num_tokens": 97057191.0,
+      "eval_runtime": 3.6724,
+      "eval_samples_per_second": 27.23,
+      "eval_steps_per_second": 3.54,
+      "step": 1300
+    },
+    {
+      "entropy": 1.4690280217677354,
+      "epoch": 2.8239973036737447,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 4.3733795402321476e-08,
+      "loss": 1.2157417297363282,
+      "mean_token_accuracy": 0.6754748554900288,
+      "num_tokens": 97819920.0,
+      "step": 1310
+    },
+    {
+      "entropy": 1.4682903721928597,
+      "epoch": 2.84556791371756,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 3.3852786882372925e-08,
+      "loss": 1.217051124572754,
+      "mean_token_accuracy": 0.6758211361244321,
+      "num_tokens": 98555979.0,
+      "step": 1320
+    },
+    {
+      "entropy": 1.4783522091805934,
+      "epoch": 2.8671385237613753,
+      "grad_norm": 0.03515625,
+      "learning_rate": 2.5227871486249168e-08,
+      "loss": 1.2258304595947265,
+      "mean_token_accuracy": 0.6732724322006106,
+      "num_tokens": 99303026.0,
+      "step": 1330
+    },
+    {
+      "entropy": 1.470925122499466,
+      "epoch": 2.8887091338051905,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 1.7863442178354117e-08,
+      "loss": 1.2253659248352051,
+      "mean_token_accuracy": 0.6761119239032268,
+      "num_tokens": 100060567.0,
+      "step": 1340
+    },
+    {
+      "entropy": 1.4869282141327858,
+      "epoch": 2.9102797438490056,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.1763249914415465e-08,
+      "loss": 1.2255492210388184,
+      "mean_token_accuracy": 0.6743376379832625,
+      "num_tokens": 100800446.0,
+      "step": 1350
+    },
+    {
+      "epoch": 2.9102797438490056,
+      "eval_entropy": 1.3712318310370812,
+      "eval_loss": 1.1809680461883545,
+      "eval_mean_token_accuracy": 0.6843105783829322,
+      "eval_num_tokens": 100800446.0,
+      "eval_runtime": 3.6629,
+      "eval_samples_per_second": 27.3,
+      "eval_steps_per_second": 3.549,
+      "step": 1350
+    },
+    {
+      "entropy": 1.4652011856436729,
+      "epoch": 2.931850353892821,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 6.930401730994574e-09,
+      "loss": 1.2218934059143067,
+      "mean_token_accuracy": 0.675122519209981,
+      "num_tokens": 101545860.0,
+      "step": 1360
+    },
+    {
+      "entropy": 1.4832888428121804,
+      "epoch": 2.9534209639366362,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 3.367359162966266e-09,
+      "loss": 1.2362324714660644,
+      "mean_token_accuracy": 0.6732010493054986,
+      "num_tokens": 102297655.0,
+      "step": 1370
+    },
+    {
+      "entropy": 1.4625216010957955,
+      "epoch": 2.9749915739804518,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.075936989775317e-09,
+      "loss": 1.2096445083618164,
+      "mean_token_accuracy": 0.6758219797164202,
+      "num_tokens": 103042675.0,
+      "step": 1380
+    },
+    {
+      "entropy": 1.511010423488915,
+      "epoch": 2.996562184024267,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 5.730231110639972e-11,
+      "loss": 1.252042579650879,
+      "mean_token_accuracy": 0.6698385804891587,
+      "num_tokens": 103799829.0,
+      "step": 1390
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1392,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2273736039996662e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/training_args.bin b/training_args.bin
new file mode 100644
index 0000000..25d793e
--- /dev/null
+++ b/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99ed9bf4d44c111b2de4411151f761d4d03895679fe864d2fe147304b2ec13b9
+size 5713