初始化项目，由ModelHub XC社区提供模型

Model: cuong1692001/Terminal-data_processing Source: Original Platform
2026-07-11 09:29:10 +08:00
commit 27971b5f78
49 changed files with 11183 additions and 0 deletions
--- a/.gitattributes
+++ b/.gitattributes
@@ -0,0 +1,37 @@
+*.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
+checkpoint-2322/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
--- a/README.md
+++ b/README.md
@@ -0,0 +1,59 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen3-8B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: Terminal-data_processing
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# Terminal-data_processing
+
+This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the nemotron_easy_data_processing, the nemotron_medium_data_processing and the nemotron_mixed_data_processing datasets.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 8
+- total_train_batch_size: 8
+- total_eval_batch_size: 64
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- num_epochs: 2.0
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.6.0
+- Pytorch 2.12.1+cu130
+- Datasets 4.0.0
+- Tokenizers 0.22.2
--- a/all_results.json
+++ b/all_results.json
@@ -0,0 +1,9 @@
+{
+    "epoch": 2.0,
+    "num_input_tokens_seen": 236749552,
+    "total_flos": 437831363461120.0,
+    "train_loss": 0.3453142216989031,
+    "train_runtime": 32003.0179,
+    "train_samples_per_second": 0.58,
+    "train_steps_per_second": 0.073
+}
--- a/chat_template.jinja
+++ b/chat_template.jinja
@@ -0,0 +1,89 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- 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' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}
--- a/checkpoint-2322/chat_template.jinja
+++ b/checkpoint-2322/chat_template.jinja
@@ -0,0 +1,89 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- 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' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}
--- a/checkpoint-2322/config.json
+++ b/checkpoint-2322/config.json
@@ -0,0 +1,71 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 12288,
+  "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",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 40960,
+  "max_window_layers": 36,
+  "model_type": "qwen3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "transformers_version": "5.6.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
--- a/checkpoint-2322/generation_config.json
+++ b/checkpoint-2322/generation_config.json
@@ -0,0 +1,12 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "5.6.0"
+}
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/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:eb0cd5f8aa364ed62e73ce85808e1ef208cbf9f0680a490be2e067234b7b3eae
+size 12286108769
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59eb31813ae19e394c2353cc1a303ee4edee140e66b736be179c9e433f4db4f5
+size 12286108769
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22c64cf0b56f7ad845dcb392cfe4b60e4577c92a7c98663c937570eea570ddf4
+size 12286108769
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/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:bb30cc99b5968c0a2c17fd8f40ef1293282aeaba60d2e916d933ebedd1409adf
+size 12286108769
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b9e8b31ee52eff391f3b11863e19711b21870f0169c21e4a062f0d90b3f4c45
+size 12286108769
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2530c8d6a4cce4aee934d6bf7ae693d542f0c1a67e5b0513f7481cf26d59eb2d
+size 12286108769
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e9edf5deafdebc4e92e88c8babd20f7a05f2ebfc1fecda4583c75a94b6ab3c9
+size 12286108769
--- a/checkpoint-2322/global_step2322/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt
+++ b/checkpoint-2322/global_step2322/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:329b2e6db419405c085b916b684a2ca4bd387d13e6569c8fde52c37eba5dfeec
+size 12286108769
--- a/checkpoint-2322/global_step2322/zero_pp_rank_0_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_0_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f897eb02dbd51f9430b74e5ac81d5107d76b3fdd0c3274660596920d4f0ead38
+size 202732
--- a/checkpoint-2322/global_step2322/zero_pp_rank_1_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_1_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0051b654a4fe656c5165eb0a5546a88d5b65cb57e02f681493e0523f25ddd584
+size 202732
--- a/checkpoint-2322/global_step2322/zero_pp_rank_2_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_2_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87236cbb047272cb7d7204e22dd1de24c0b6fad111ab8ce195c910f988a37453
+size 202732
--- a/checkpoint-2322/global_step2322/zero_pp_rank_3_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_3_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe0017adb39f035f7fe51d8b8d6927fbf3272a191b18db9e449b38734564b657
+size 202732
--- a/checkpoint-2322/global_step2322/zero_pp_rank_4_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_4_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3196f105ccfe8eefefbadfd60b13d450d6d6495bf5608245d23147d524343e87
+size 202732
--- a/checkpoint-2322/global_step2322/zero_pp_rank_5_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_5_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:398ca84fe5680e9a02ac90f757b5ea15c66d838ae65e385658299f9dfa08cbe6
+size 202732
--- a/checkpoint-2322/global_step2322/zero_pp_rank_6_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_6_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a5a4e0eed56043712232b23f0007572c92debb7d5287a2d203c1aeaee496bb5
+size 202732
--- a/checkpoint-2322/global_step2322/zero_pp_rank_7_mp_rank_00_model_states.pt
+++ b/checkpoint-2322/global_step2322/zero_pp_rank_7_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a045da4a179ce8fa8cf668705dc8bf62895638b26f29333bef7134ea60c3830e
+size 202732
--- a/checkpoint-2322/latest
+++ b/checkpoint-2322/latest
@@ -0,0 +1 @@
+global_step2322
--- a/checkpoint-2322/model.safetensors
+++ b/checkpoint-2322/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:044a4204bb4018999a53272406568d66294e6298cebaa0dcfb55f5cd4908b269
+size 16381517208
--- a/checkpoint-2322/rng_state_0.pth
+++ b/checkpoint-2322/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bfe1981024ef92f2da08a90c72c7c793d1cc9de1547abd2556c968be70232eb
+size 16389
--- a/checkpoint-2322/rng_state_1.pth
+++ b/checkpoint-2322/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a35b845d476d830805793c3dcf8ac2daad87fec289bff3f7eda9e72fc374eda1
+size 16389
--- a/checkpoint-2322/rng_state_2.pth
+++ b/checkpoint-2322/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03e9880996b01262a807d1ec3ebd91eee540e08130a14a45a4648731fd0d48a9
+size 16389
--- a/checkpoint-2322/rng_state_3.pth
+++ b/checkpoint-2322/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee25c237d6fe62ec76adcf7daf899d7ed32eab5d1a5b447b911f4451c9a1b258
+size 16389
--- a/checkpoint-2322/rng_state_4.pth
+++ b/checkpoint-2322/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a6b31133f29a8fc0cb538aa807d6a403bd51939336bfd425cd3d122d8c5595c
+size 16389
--- a/checkpoint-2322/rng_state_5.pth
+++ b/checkpoint-2322/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a26c55b5c7fa0522b1d27b2c00a7ea77ad010f19a1321991165c5c972b8fa97a
+size 16389
--- a/checkpoint-2322/rng_state_6.pth
+++ b/checkpoint-2322/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a1a3cf85626196804f25a8293e22dc561bba068a70fb123e04afe4896c33972
+size 16389
--- a/checkpoint-2322/rng_state_7.pth
+++ b/checkpoint-2322/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28f87c1ee5f5db346c7b913137cbccd196eaf8ec5a4cf9f192418a3069269b49
+size 16389
--- a/checkpoint-2322/scheduler.pt
+++ b/checkpoint-2322/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d4c952a366e2e1ca694ca0f71cdc468c9f445139fe6e906c1076c4860074851
+size 1465
--- a/checkpoint-2322/tokenizer.json
+++ b/checkpoint-2322/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650
--- a/checkpoint-2322/tokenizer_config.json
+++ b/checkpoint-2322/tokenizer_config.json
@@ -0,0 +1,31 @@
+{
+  "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,
+  "local_files_only": false,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
--- a/checkpoint-2322/trainer_state.json
+++ b/checkpoint-2322/trainer_state.json
--- a/checkpoint-2322/training_args.bin
+++ b/checkpoint-2322/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e1e995bcc45bb5d176ef43a23610ceefbca95a9495933ad7c576c2d32115034
+size 7697
--- a/checkpoint-2322/zero_to_fp32.py
+++ b/checkpoint-2322/zero_to_fp32.py
@@ -0,0 +1,760 @@
+#!/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 . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+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, weights_only=False)
+
+        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 tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # 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 ZERO_STAGE not 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}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] 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, exclude_frozen_parameters):
+    """
+    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,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+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 _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+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() if _has_callable(shape, 'numel') else math.prod(shape)
+            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,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _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")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+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 = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * 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
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        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}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        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,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _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 to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    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``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    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.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    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")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    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_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``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``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+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("Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info("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_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    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("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    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_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)
--- a/config.json
+++ b/config.json
@@ -0,0 +1,71 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 12288,
+  "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",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 40960,
+  "max_window_layers": 36,
+  "model_type": "qwen3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "transformers_version": "5.6.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
--- a/generation_config.json
+++ b/generation_config.json
@@ -0,0 +1,12 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "5.6.0"
+}
--- a/model.safetensors
+++ b/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:044a4204bb4018999a53272406568d66294e6298cebaa0dcfb55f5cd4908b269
+size 16381517208
--- a/tokenizer.json
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650
--- a/tokenizer_config.json
+++ b/tokenizer_config.json
@@ -0,0 +1,17 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
--- a/train_results.json
+++ b/train_results.json
@@ -0,0 +1,9 @@
+{
+    "epoch": 2.0,
+    "num_input_tokens_seen": 236749552,
+    "total_flos": 437831363461120.0,
+    "train_loss": 0.3453142216989031,
+    "train_runtime": 32003.0179,
+    "train_samples_per_second": 0.58,
+    "train_steps_per_second": 0.073
+}
--- a/trainer_log.jsonl
+++ b/trainer_log.jsonl
@@ -0,0 +1,465 @@
+{"current_steps": 5, "total_steps": 2322, "loss": 0.7518440246582031, "lr": 9.999926779237713e-06, "epoch": 0.004306632213608958, "percentage": 0.22, "elapsed_time": "0:01:20", "remaining_time": "10:20:13", "throughput": 6431.2, "total_tokens": 516464}
+{"current_steps": 10, "total_steps": 2322, "loss": 0.5895246028900146, "lr": 9.999629323566323e-06, "epoch": 0.008613264427217916, "percentage": 0.43, "elapsed_time": "0:02:36", "remaining_time": "10:01:27", "throughput": 6733.38, "total_tokens": 1050992}
+{"current_steps": 15, "total_steps": 2322, "loss": 0.5217817306518555, "lr": 9.999103070290155e-06, "epoch": 0.012919896640826873, "percentage": 0.65, "elapsed_time": "0:03:41", "remaining_time": "9:27:41", "throughput": 7185.92, "total_tokens": 1591440}
+{"current_steps": 20, "total_steps": 2322, "loss": 0.5254570960998535, "lr": 9.998348043492093e-06, "epoch": 0.017226528854435832, "percentage": 0.86, "elapsed_time": "0:04:41", "remaining_time": "8:59:04", "throughput": 7488.43, "total_tokens": 2104328}
+{"current_steps": 25, "total_steps": 2322, "loss": 0.49294347763061525, "lr": 9.997364277724362e-06, "epoch": 0.02153316106804479, "percentage": 1.08, "elapsed_time": "0:05:42", "remaining_time": "8:44:04", "throughput": 7770.67, "total_tokens": 2659352}
+{"current_steps": 30, "total_steps": 2322, "loss": 0.4858407020568848, "lr": 9.996151818006951e-06, "epoch": 0.025839793281653745, "percentage": 1.29, "elapsed_time": "0:06:51", "remaining_time": "8:43:28", "throughput": 7782.82, "total_tokens": 3199536}
+{"current_steps": 35, "total_steps": 2322, "loss": 0.4974493980407715, "lr": 9.99471071982555e-06, "epoch": 0.030146425495262703, "percentage": 1.51, "elapsed_time": "0:07:57", "remaining_time": "8:39:56", "throughput": 7677.56, "total_tokens": 3665456}
+{"current_steps": 40, "total_steps": 2322, "loss": 0.4639857292175293, "lr": 9.993041049129005e-06, "epoch": 0.034453057708871665, "percentage": 1.72, "elapsed_time": "0:09:08", "remaining_time": "8:41:40", "throughput": 7621.99, "total_tokens": 4181744}
+{"current_steps": 45, "total_steps": 2322, "loss": 0.4471879482269287, "lr": 9.99114288232632e-06, "epoch": 0.03875968992248062, "percentage": 1.94, "elapsed_time": "0:10:25", "remaining_time": "8:47:37", "throughput": 7552.78, "total_tokens": 4725304}
+{"current_steps": 50, "total_steps": 2322, "loss": 0.4194998264312744, "lr": 9.989016306283132e-06, "epoch": 0.04306632213608958, "percentage": 2.15, "elapsed_time": "0:11:35", "remaining_time": "8:46:29", "throughput": 7505.68, "total_tokens": 5217928}
+{"current_steps": 55, "total_steps": 2322, "loss": 0.4558732509613037, "lr": 9.986661418317759e-06, "epoch": 0.047372954349698536, "percentage": 2.37, "elapsed_time": "0:12:42", "remaining_time": "8:43:58", "throughput": 7499.57, "total_tokens": 5720264}
+{"current_steps": 60, "total_steps": 2322, "loss": 0.4496774673461914, "lr": 9.984078326196736e-06, "epoch": 0.05167958656330749, "percentage": 2.58, "elapsed_time": "0:13:49", "remaining_time": "8:41:29", "throughput": 7471.22, "total_tokens": 6200784}
+{"current_steps": 65, "total_steps": 2322, "loss": 0.4551366329193115, "lr": 9.981267148129884e-06, "epoch": 0.05598621877691645, "percentage": 2.8, "elapsed_time": "0:15:01", "remaining_time": "8:41:55", "throughput": 7452.45, "total_tokens": 6721120}
+{"current_steps": 70, "total_steps": 2322, "loss": 0.43771700859069823, "lr": 9.978228012764904e-06, "epoch": 0.06029285099052541, "percentage": 3.01, "elapsed_time": "0:16:05", "remaining_time": "8:37:32", "throughput": 7461.71, "total_tokens": 7202096}
+{"current_steps": 75, "total_steps": 2322, "loss": 0.44599485397338867, "lr": 9.974961059181482e-06, "epoch": 0.06459948320413436, "percentage": 3.23, "elapsed_time": "0:17:15", "remaining_time": "8:36:54", "throughput": 7459.1, "total_tokens": 7721720}
+{"current_steps": 80, "total_steps": 2322, "loss": 0.4492579460144043, "lr": 9.971466436884933e-06, "epoch": 0.06890611541774333, "percentage": 3.45, "elapsed_time": "0:18:20", "remaining_time": "8:34:07", "throughput": 7463.88, "total_tokens": 8215480}
+{"current_steps": 85, "total_steps": 2322, "loss": 0.428343677520752, "lr": 9.967744305799358e-06, "epoch": 0.07321274763135228, "percentage": 3.66, "elapsed_time": "0:19:35", "remaining_time": "8:35:28", "throughput": 7446.94, "total_tokens": 8751608}
+{"current_steps": 90, "total_steps": 2322, "loss": 0.44635858535766604, "lr": 9.963794836260316e-06, "epoch": 0.07751937984496124, "percentage": 3.88, "elapsed_time": "0:20:49", "remaining_time": "8:36:16", "throughput": 7381.8, "total_tokens": 9220400}
+{"current_steps": 95, "total_steps": 2322, "loss": 0.41611676216125487, "lr": 9.959618209007045e-06, "epoch": 0.0818260120585702, "percentage": 4.09, "elapsed_time": "0:22:08", "remaining_time": "8:39:11", "throughput": 7339.04, "total_tokens": 9752480}
+{"current_steps": 100, "total_steps": 2322, "loss": 0.4301795959472656, "lr": 9.955214615174174e-06, "epoch": 0.08613264427217916, "percentage": 4.31, "elapsed_time": "0:23:17", "remaining_time": "8:37:39", "throughput": 7358.5, "total_tokens": 10286000}
+{"current_steps": 105, "total_steps": 2322, "loss": 0.42357568740844725, "lr": 9.950584256282988e-06, "epoch": 0.09043927648578812, "percentage": 4.52, "elapsed_time": "0:24:25", "remaining_time": "8:35:38", "throughput": 7361.76, "total_tokens": 10786992}
+{"current_steps": 110, "total_steps": 2322, "loss": 0.43195395469665526, "lr": 9.945727344232202e-06, "epoch": 0.09474590869939707, "percentage": 4.74, "elapsed_time": "0:25:43", "remaining_time": "8:37:28", "throughput": 7334.76, "total_tokens": 11324728}
+{"current_steps": 115, "total_steps": 2322, "loss": 0.4098663806915283, "lr": 9.940644101288259e-06, "epoch": 0.09905254091300603, "percentage": 4.95, "elapsed_time": "0:26:50", "remaining_time": "8:35:00", "throughput": 7354.92, "total_tokens": 11842384}
+{"current_steps": 120, "total_steps": 2322, "loss": 0.4260974884033203, "lr": 9.93533476007517e-06, "epoch": 0.10335917312661498, "percentage": 5.17, "elapsed_time": "0:27:59", "remaining_time": "8:33:41", "throughput": 7379.61, "total_tokens": 12395312}
+{"current_steps": 125, "total_steps": 2322, "loss": 0.41834230422973634, "lr": 9.929799563563858e-06, "epoch": 0.10766580534022395, "percentage": 5.38, "elapsed_time": "0:29:08", "remaining_time": "8:32:10", "throughput": 7364.63, "total_tokens": 12876440}
+{"current_steps": 130, "total_steps": 2322, "loss": 0.4106014251708984, "lr": 9.924038765061042e-06, "epoch": 0.1119724375538329, "percentage": 5.6, "elapsed_time": "0:30:18", "remaining_time": "8:30:59", "throughput": 7356.64, "total_tokens": 13376496}
+{"current_steps": 135, "total_steps": 2322, "loss": 0.4307220458984375, "lr": 9.918052628197645e-06, "epoch": 0.11627906976744186, "percentage": 5.81, "elapsed_time": "0:31:19", "remaining_time": "8:27:19", "throughput": 7382.77, "total_tokens": 13872280}
+{"current_steps": 140, "total_steps": 2322, "loss": 0.4186269760131836, "lr": 9.911841426916733e-06, "epoch": 0.12058570198105081, "percentage": 6.03, "elapsed_time": "0:32:28", "remaining_time": "8:26:16", "throughput": 7374.92, "total_tokens": 14373496}
+{"current_steps": 145, "total_steps": 2322, "loss": 0.41518564224243165, "lr": 9.905405445460972e-06, "epoch": 0.12489233419465978, "percentage": 6.24, "elapsed_time": "0:33:33", "remaining_time": "8:23:56", "throughput": 7391.73, "total_tokens": 14886552}
+{"current_steps": 150, "total_steps": 2322, "loss": 0.40087127685546875, "lr": 9.898744978359627e-06, "epoch": 0.12919896640826872, "percentage": 6.46, "elapsed_time": "0:34:40", "remaining_time": "8:22:07", "throughput": 7384.67, "total_tokens": 15364928}
+{"current_steps": 155, "total_steps": 2322, "loss": 0.4161245346069336, "lr": 9.891860330415077e-06, "epoch": 0.1335055986218777, "percentage": 6.68, "elapsed_time": "0:35:48", "remaining_time": "8:20:32", "throughput": 7391.17, "total_tokens": 15877064}
+{"current_steps": 160, "total_steps": 2322, "loss": 0.38726229667663575, "lr": 9.884751816688873e-06, "epoch": 0.13781223083548666, "percentage": 6.89, "elapsed_time": "0:36:59", "remaining_time": "8:19:47", "throughput": 7372.99, "total_tokens": 16362424}
+{"current_steps": 165, "total_steps": 2322, "loss": 0.41555061340332033, "lr": 9.877419762487312e-06, "epoch": 0.1421188630490956, "percentage": 7.11, "elapsed_time": "0:38:05", "remaining_time": "8:17:51", "throughput": 7389.11, "total_tokens": 16884272}
+{"current_steps": 170, "total_steps": 2322, "loss": 0.42660012245178225, "lr": 9.869864503346557e-06, "epoch": 0.14642549526270457, "percentage": 7.32, "elapsed_time": "0:39:13", "remaining_time": "8:16:33", "throughput": 7391.17, "total_tokens": 17395904}
+{"current_steps": 175, "total_steps": 2322, "loss": 0.3985039472579956, "lr": 9.862086385017283e-06, "epoch": 0.1507321274763135, "percentage": 7.54, "elapsed_time": "0:40:25", "remaining_time": "8:15:58", "throughput": 7394.18, "total_tokens": 17935128}
+{"current_steps": 180, "total_steps": 2322, "loss": 0.403258228302002, "lr": 9.854085763448844e-06, "epoch": 0.15503875968992248, "percentage": 7.75, "elapsed_time": "0:41:31", "remaining_time": "8:14:05", "throughput": 7402.5, "total_tokens": 18441240}
+{"current_steps": 185, "total_steps": 2322, "loss": 0.4074747085571289, "lr": 9.845863004772994e-06, "epoch": 0.15934539190353145, "percentage": 7.97, "elapsed_time": "0:42:42", "remaining_time": "8:13:19", "throughput": 7395.48, "total_tokens": 18950248}
+{"current_steps": 190, "total_steps": 2322, "loss": 0.3952253580093384, "lr": 9.837418485287126e-06, "epoch": 0.1636520241171404, "percentage": 8.18, "elapsed_time": "0:43:49", "remaining_time": "8:11:43", "throughput": 7394.8, "total_tokens": 19443272}
+{"current_steps": 195, "total_steps": 2322, "loss": 0.4341114521026611, "lr": 9.82875259143706e-06, "epoch": 0.16795865633074936, "percentage": 8.4, "elapsed_time": "0:45:01", "remaining_time": "8:11:10", "throughput": 7392.86, "total_tokens": 19974056}
+{"current_steps": 200, "total_steps": 2322, "loss": 0.39603476524353026, "lr": 9.819865719799344e-06, "epoch": 0.17226528854435832, "percentage": 8.61, "elapsed_time": "0:46:00", "remaining_time": "8:08:12", "throughput": 7427.14, "total_tokens": 20505304}
+{"current_steps": 205, "total_steps": 2322, "loss": 0.3957651615142822, "lr": 9.81075827706312e-06, "epoch": 0.17657192075796727, "percentage": 8.83, "elapsed_time": "0:47:13", "remaining_time": "8:07:36", "throughput": 7427.03, "total_tokens": 21041488}
+{"current_steps": 210, "total_steps": 2322, "loss": 0.3961822509765625, "lr": 9.801430680011499e-06, "epoch": 0.18087855297157623, "percentage": 9.04, "elapsed_time": "0:48:21", "remaining_time": "8:06:25", "throughput": 7417.56, "total_tokens": 21525384}
+{"current_steps": 215, "total_steps": 2322, "loss": 0.42037005424499513, "lr": 9.791883355502503e-06, "epoch": 0.18518518518518517, "percentage": 9.26, "elapsed_time": "0:49:32", "remaining_time": "8:05:28", "throughput": 7406.98, "total_tokens": 22015656}
+{"current_steps": 220, "total_steps": 2322, "loss": 0.3965612888336182, "lr": 9.782116740449515e-06, "epoch": 0.18949181739879414, "percentage": 9.47, "elapsed_time": "0:50:44", "remaining_time": "8:04:46", "throughput": 7400.83, "total_tokens": 22529696}
+{"current_steps": 225, "total_steps": 2322, "loss": 0.44352130889892577, "lr": 9.7721312818013e-06, "epoch": 0.1937984496124031, "percentage": 9.69, "elapsed_time": "0:51:56", "remaining_time": "8:04:02", "throughput": 7393.26, "total_tokens": 23038360}
+{"current_steps": 230, "total_steps": 2322, "loss": 0.38213863372802737, "lr": 9.761927436521534e-06, "epoch": 0.19810508182601205, "percentage": 9.91, "elapsed_time": "0:52:59", "remaining_time": "8:01:59", "throughput": 7401.5, "total_tokens": 23532864}
+{"current_steps": 235, "total_steps": 2322, "loss": 0.3937716484069824, "lr": 9.751505671567914e-06, "epoch": 0.20241171403962102, "percentage": 10.12, "elapsed_time": "0:54:13", "remaining_time": "8:01:38", "throughput": 7377.43, "total_tokens": 24005936}
+{"current_steps": 240, "total_steps": 2322, "loss": 0.40571026802062987, "lr": 9.740866463870762e-06, "epoch": 0.20671834625322996, "percentage": 10.34, "elapsed_time": "0:55:22", "remaining_time": "8:00:26", "throughput": 7370.15, "total_tokens": 24490872}
+{"current_steps": 245, "total_steps": 2322, "loss": 0.411739444732666, "lr": 9.730010300311226e-06, "epoch": 0.21102497846683893, "percentage": 10.55, "elapsed_time": "0:56:32", "remaining_time": "7:59:16", "throughput": 7374.41, "total_tokens": 25014232}
+{"current_steps": 250, "total_steps": 2322, "loss": 0.40146331787109374, "lr": 9.718937677698976e-06, "epoch": 0.2153316106804479, "percentage": 10.77, "elapsed_time": "0:57:37", "remaining_time": "7:57:35", "throughput": 7377.39, "total_tokens": 25507448}
+{"current_steps": 255, "total_steps": 2322, "loss": 0.3829845428466797, "lr": 9.707649102749488e-06, "epoch": 0.21963824289405684, "percentage": 10.98, "elapsed_time": "0:58:46", "remaining_time": "7:56:21", "throughput": 7374.8, "total_tokens": 26003584}
+{"current_steps": 260, "total_steps": 2322, "loss": 0.37958288192749023, "lr": 9.696145092060843e-06, "epoch": 0.2239448751076658, "percentage": 11.2, "elapsed_time": "0:59:46", "remaining_time": "7:54:04", "throughput": 7395.23, "total_tokens": 26523584}
+{"current_steps": 265, "total_steps": 2322, "loss": 0.4073479652404785, "lr": 9.684426172090084e-06, "epoch": 0.22825150732127478, "percentage": 11.41, "elapsed_time": "1:01:03", "remaining_time": "7:53:59", "throughput": 7369.88, "total_tokens": 27002040}
+{"current_steps": 270, "total_steps": 2322, "loss": 0.40698766708374023, "lr": 9.672492879129136e-06, "epoch": 0.23255813953488372, "percentage": 11.63, "elapsed_time": "1:02:15", "remaining_time": "7:53:11", "throughput": 7362.62, "total_tokens": 27505176}
+{"current_steps": 275, "total_steps": 2322, "loss": 0.4029866695404053, "lr": 9.660345759280254e-06, "epoch": 0.2368647717484927, "percentage": 11.84, "elapsed_time": "1:03:21", "remaining_time": "7:51:35", "throughput": 7372.0, "total_tokens": 28022920}
+{"current_steps": 280, "total_steps": 2322, "loss": 0.4014443397521973, "lr": 9.647985368431031e-06, "epoch": 0.24117140396210163, "percentage": 12.06, "elapsed_time": "1:04:39", "remaining_time": "7:51:33", "throughput": 7354.97, "total_tokens": 28534944}
+{"current_steps": 285, "total_steps": 2322, "loss": 0.4226128101348877, "lr": 9.63541227222897e-06, "epoch": 0.2454780361757106, "percentage": 12.27, "elapsed_time": "1:05:48", "remaining_time": "7:50:17", "throughput": 7354.41, "total_tokens": 29035280}
+{"current_steps": 290, "total_steps": 2322, "loss": 0.4024317264556885, "lr": 9.622627046055584e-06, "epoch": 0.24978466838931956, "percentage": 12.49, "elapsed_time": "1:07:09", "remaining_time": "7:50:31", "throughput": 7345.84, "total_tokens": 29597328}
+{"current_steps": 295, "total_steps": 2322, "loss": 0.3940183162689209, "lr": 9.609630275000072e-06, "epoch": 0.2540913006029285, "percentage": 12.7, "elapsed_time": "1:08:20", "remaining_time": "7:49:32", "throughput": 7340.1, "total_tokens": 30094560}
+{"current_steps": 300, "total_steps": 2322, "loss": 0.3797435760498047, "lr": 9.596422553832547e-06, "epoch": 0.25839793281653745, "percentage": 12.92, "elapsed_time": "1:09:31", "remaining_time": "7:48:34", "throughput": 7342.59, "total_tokens": 30628392}
+{"current_steps": 305, "total_steps": 2322, "loss": 0.3960827350616455, "lr": 9.583004486976813e-06, "epoch": 0.26270456503014644, "percentage": 13.14, "elapsed_time": "1:10:39", "remaining_time": "7:47:17", "throughput": 7348.48, "total_tokens": 31155408}
+{"current_steps": 310, "total_steps": 2322, "loss": 0.38438570499420166, "lr": 9.5693766884827e-06, "epoch": 0.2670111972437554, "percentage": 13.35, "elapsed_time": "1:11:44", "remaining_time": "7:45:36", "throughput": 7352.66, "total_tokens": 31648472}
+{"current_steps": 315, "total_steps": 2322, "loss": 0.4010166645050049, "lr": 9.555539781997978e-06, "epoch": 0.2713178294573643, "percentage": 13.57, "elapsed_time": "1:12:55", "remaining_time": "7:44:35", "throughput": 7348.16, "total_tokens": 32148440}
+{"current_steps": 320, "total_steps": 2322, "loss": 0.3938159465789795, "lr": 9.541494400739799e-06, "epoch": 0.2756244616709733, "percentage": 13.78, "elapsed_time": "1:14:09", "remaining_time": "7:43:56", "throughput": 7340.16, "total_tokens": 32659456}
+{"current_steps": 325, "total_steps": 2322, "loss": 0.39118094444274903, "lr": 9.527241187465735e-06, "epoch": 0.27993109388458226, "percentage": 14.0, "elapsed_time": "1:15:14", "remaining_time": "7:42:22", "throughput": 7344.1, "total_tokens": 33157552}
+{"current_steps": 330, "total_steps": 2322, "loss": 0.38675827980041505, "lr": 9.51278079444435e-06, "epoch": 0.2842377260981912, "percentage": 14.21, "elapsed_time": "1:16:19", "remaining_time": "7:40:42", "throughput": 7356.42, "total_tokens": 33687624}
+{"current_steps": 335, "total_steps": 2322, "loss": 0.4036139488220215, "lr": 9.498113883425364e-06, "epoch": 0.2885443583118002, "percentage": 14.43, "elapsed_time": "1:17:33", "remaining_time": "7:40:01", "throughput": 7343.18, "total_tokens": 34171792}
+{"current_steps": 340, "total_steps": 2322, "loss": 0.41157045364379885, "lr": 9.483241125609358e-06, "epoch": 0.29285099052540914, "percentage": 14.64, "elapsed_time": "1:18:49", "remaining_time": "7:39:31", "throughput": 7334.7, "total_tokens": 34691536}
+{"current_steps": 345, "total_steps": 2322, "loss": 0.42139220237731934, "lr": 9.468163201617063e-06, "epoch": 0.2971576227390181, "percentage": 14.86, "elapsed_time": "1:20:01", "remaining_time": "7:38:32", "throughput": 7333.03, "total_tokens": 35206192}
+{"current_steps": 350, "total_steps": 2322, "loss": 0.3920948028564453, "lr": 9.45288080145821e-06, "epoch": 0.301464254952627, "percentage": 15.07, "elapsed_time": "1:21:13", "remaining_time": "7:37:40", "throughput": 7326.03, "total_tokens": 35705960}
+{"current_steps": 355, "total_steps": 2322, "loss": 0.37868556976318357, "lr": 9.437394624499957e-06, "epoch": 0.305770887166236, "percentage": 15.29, "elapsed_time": "1:22:14", "remaining_time": "7:35:43", "throughput": 7337.03, "total_tokens": 36206960}
+{"current_steps": 360, "total_steps": 2322, "loss": 0.3680633544921875, "lr": 9.421705379434887e-06, "epoch": 0.31007751937984496, "percentage": 15.5, "elapsed_time": "1:23:29", "remaining_time": "7:34:59", "throughput": 7336.5, "total_tokens": 36749336}
+{"current_steps": 365, "total_steps": 2322, "loss": 0.386813759803772, "lr": 9.405813784248562e-06, "epoch": 0.3143841515934539, "percentage": 15.72, "elapsed_time": "1:24:28", "remaining_time": "7:32:58", "throughput": 7349.67, "total_tokens": 37255480}
+{"current_steps": 370, "total_steps": 2322, "loss": 0.38484036922454834, "lr": 9.38972056618668e-06, "epoch": 0.3186907838070629, "percentage": 15.93, "elapsed_time": "1:25:30", "remaining_time": "7:31:09", "throughput": 7359.22, "total_tokens": 37760024}
+{"current_steps": 375, "total_steps": 2322, "loss": 0.3900888919830322, "lr": 9.37342646172179e-06, "epoch": 0.32299741602067183, "percentage": 16.15, "elapsed_time": "1:26:38", "remaining_time": "7:29:48", "throughput": 7370.92, "total_tokens": 38314528}
+{"current_steps": 380, "total_steps": 2322, "loss": 0.3763136625289917, "lr": 9.356932216519588e-06, "epoch": 0.3273040482342808, "percentage": 16.37, "elapsed_time": "1:27:40", "remaining_time": "7:28:04", "throughput": 7380.68, "total_tokens": 38827464}
+{"current_steps": 385, "total_steps": 2322, "loss": 0.38970165252685546, "lr": 9.340238585404787e-06, "epoch": 0.33161068044788977, "percentage": 16.58, "elapsed_time": "1:28:54", "remaining_time": "7:27:19", "throughput": 7371.87, "total_tokens": 39325920}
+{"current_steps": 390, "total_steps": 2322, "loss": 0.3639005184173584, "lr": 9.32334633232659e-06, "epoch": 0.3359173126614987, "percentage": 16.8, "elapsed_time": "1:30:04", "remaining_time": "7:26:12", "throughput": 7370.51, "total_tokens": 39832488}
+{"current_steps": 395, "total_steps": 2322, "loss": 0.3740285873413086, "lr": 9.306256230323714e-06, "epoch": 0.34022394487510765, "percentage": 17.01, "elapsed_time": "1:31:15", "remaining_time": "7:25:10", "throughput": 7370.94, "total_tokens": 40357880}
+{"current_steps": 400, "total_steps": 2322, "loss": 0.39269113540649414, "lr": 9.28896906148902e-06, "epoch": 0.34453057708871665, "percentage": 17.23, "elapsed_time": "1:32:32", "remaining_time": "7:24:40", "throughput": 7363.83, "total_tokens": 40888088}
+{"current_steps": 405, "total_steps": 2322, "loss": 0.38478426933288573, "lr": 9.271485616933725e-06, "epoch": 0.3488372093023256, "percentage": 17.44, "elapsed_time": "1:33:42", "remaining_time": "7:23:35", "throughput": 7362.02, "total_tokens": 41396320}
+{"current_steps": 410, "total_steps": 2322, "loss": 0.38239054679870604, "lr": 9.253806696751185e-06, "epoch": 0.35314384151593453, "percentage": 17.66, "elapsed_time": "1:34:49", "remaining_time": "7:22:10", "throughput": 7362.02, "total_tokens": 41883152}
+{"current_steps": 415, "total_steps": 2322, "loss": 0.39941422939300536, "lr": 9.235933109980302e-06, "epoch": 0.35745047372954347, "percentage": 17.87, "elapsed_time": "1:36:03", "remaining_time": "7:21:24", "throughput": 7354.35, "total_tokens": 42386288}
+{"current_steps": 420, "total_steps": 2322, "loss": 0.3732919216156006, "lr": 9.217865674568482e-06, "epoch": 0.36175710594315247, "percentage": 18.09, "elapsed_time": "1:37:10", "remaining_time": "7:20:05", "throughput": 7360.95, "total_tokens": 42921232}
+{"current_steps": 425, "total_steps": 2322, "loss": 0.36808338165283205, "lr": 9.19960521733421e-06, "epoch": 0.3660637381567614, "percentage": 18.3, "elapsed_time": "1:38:19", "remaining_time": "7:18:52", "throughput": 7361.87, "total_tokens": 43431736}
+{"current_steps": 430, "total_steps": 2322, "loss": 0.3811118841171265, "lr": 9.181152573929215e-06, "epoch": 0.37037037037037035, "percentage": 18.52, "elapsed_time": "1:39:27", "remaining_time": "7:17:38", "throughput": 7363.37, "total_tokens": 43944248}
+{"current_steps": 435, "total_steps": 2322, "loss": 0.40485553741455077, "lr": 9.162508588800221e-06, "epoch": 0.37467700258397935, "percentage": 18.73, "elapsed_time": "1:40:44", "remaining_time": "7:17:02", "throughput": 7356.63, "total_tokens": 44470472}
+{"current_steps": 440, "total_steps": 2322, "loss": 0.3950349807739258, "lr": 9.143674115150306e-06, "epoch": 0.3789836347975883, "percentage": 18.95, "elapsed_time": "1:41:55", "remaining_time": "7:15:57", "throughput": 7358.03, "total_tokens": 44997504}
+{"current_steps": 445, "total_steps": 2322, "loss": 0.3851844310760498, "lr": 9.124650014899868e-06, "epoch": 0.3832902670111972, "percentage": 19.16, "elapsed_time": "1:43:03", "remaining_time": "7:14:43", "throughput": 7356.21, "total_tokens": 45490008}
+{"current_steps": 450, "total_steps": 2322, "loss": 0.40310277938842776, "lr": 9.105437158647154e-06, "epoch": 0.3875968992248062, "percentage": 19.38, "elapsed_time": "1:44:03", "remaining_time": "7:12:53", "throughput": 7372.07, "total_tokens": 46027896}
+{"current_steps": 455, "total_steps": 2322, "loss": 0.39217004776000974, "lr": 9.086036425628453e-06, "epoch": 0.39190353143841516, "percentage": 19.6, "elapsed_time": "1:45:13", "remaining_time": "7:11:45", "throughput": 7372.15, "total_tokens": 46543128}
+{"current_steps": 460, "total_steps": 2322, "loss": 0.3778968811035156, "lr": 9.066448703677828e-06, "epoch": 0.3962101636520241, "percentage": 19.81, "elapsed_time": "1:46:22", "remaining_time": "7:10:36", "throughput": 7371.09, "total_tokens": 47048496}
+{"current_steps": 465, "total_steps": 2322, "loss": 0.3833005905151367, "lr": 9.046674889186509e-06, "epoch": 0.4005167958656331, "percentage": 20.03, "elapsed_time": "1:47:39", "remaining_time": "7:09:56", "throughput": 7366.2, "total_tokens": 47582216}
+{"current_steps": 470, "total_steps": 2322, "loss": 0.38653390407562255, "lr": 9.026715887061858e-06, "epoch": 0.40482342807924204, "percentage": 20.24, "elapsed_time": "1:48:51", "remaining_time": "7:08:56", "throughput": 7365.0, "total_tokens": 48103904}
+{"current_steps": 475, "total_steps": 2322, "loss": 0.3771857738494873, "lr": 9.006572610685969e-06, "epoch": 0.409130060292851, "percentage": 20.46, "elapsed_time": "1:49:59", "remaining_time": "7:07:39", "throughput": 7363.18, "total_tokens": 48589720}
+{"current_steps": 480, "total_steps": 2322, "loss": 0.39016103744506836, "lr": 8.986245981873854e-06, "epoch": 0.4134366925064599, "percentage": 20.67, "elapsed_time": "1:51:09", "remaining_time": "7:06:34", "throughput": 7364.73, "total_tokens": 49120400}
+{"current_steps": 485, "total_steps": 2322, "loss": 0.398733115196228, "lr": 8.965736930831272e-06, "epoch": 0.4177433247200689, "percentage": 20.89, "elapsed_time": "1:52:20", "remaining_time": "7:05:31", "throughput": 7365.82, "total_tokens": 49651912}
+{"current_steps": 490, "total_steps": 2322, "loss": 0.3702796697616577, "lr": 8.945046396112158e-06, "epoch": 0.42204995693367786, "percentage": 21.1, "elapsed_time": "1:53:39", "remaining_time": "7:04:58", "throughput": 7355.57, "total_tokens": 50164680}
+{"current_steps": 495, "total_steps": 2322, "loss": 0.36701819896697996, "lr": 8.92417532457566e-06, "epoch": 0.4263565891472868, "percentage": 21.32, "elapsed_time": "1:54:43", "remaining_time": "7:03:26", "throughput": 7361.63, "total_tokens": 50674248}
+{"current_steps": 500, "total_steps": 2322, "loss": 0.3911170959472656, "lr": 8.903124671342822e-06, "epoch": 0.4306632213608958, "percentage": 21.53, "elapsed_time": "1:55:55", "remaining_time": "7:02:24", "throughput": 7361.41, "total_tokens": 51200088}
+{"current_steps": 505, "total_steps": 2322, "loss": 0.39371590614318847, "lr": 8.881895399752873e-06, "epoch": 0.43496985357450474, "percentage": 21.75, "elapsed_time": "1:57:04", "remaining_time": "7:01:13", "throughput": 7366.1, "total_tokens": 51742280}
+{"current_steps": 510, "total_steps": 2322, "loss": 0.3914726972579956, "lr": 8.86048848131913e-06, "epoch": 0.4392764857881137, "percentage": 21.96, "elapsed_time": "1:58:12", "remaining_time": "6:59:57", "throughput": 7373.54, "total_tokens": 52294208}
+{"current_steps": 515, "total_steps": 2322, "loss": 0.39632158279418944, "lr": 8.838904895684555e-06, "epoch": 0.4435831180017227, "percentage": 22.18, "elapsed_time": "1:59:27", "remaining_time": "6:59:09", "throughput": 7370.86, "total_tokens": 52831024}
+{"current_steps": 520, "total_steps": 2322, "loss": 0.38961071968078614, "lr": 8.81714563057691e-06, "epoch": 0.4478897502153316, "percentage": 22.39, "elapsed_time": "2:00:38", "remaining_time": "6:58:03", "throughput": 7373.76, "total_tokens": 53372888}
+{"current_steps": 525, "total_steps": 2322, "loss": 0.37157084941864016, "lr": 8.795211681763565e-06, "epoch": 0.45219638242894056, "percentage": 22.61, "elapsed_time": "2:01:49", "remaining_time": "6:56:58", "throughput": 7370.08, "total_tokens": 53869408}
+{"current_steps": 530, "total_steps": 2322, "loss": 0.37537417411804197, "lr": 8.773104053005928e-06, "epoch": 0.45650301464254955, "percentage": 22.83, "elapsed_time": "2:03:00", "remaining_time": "6:55:53", "throughput": 7366.9, "total_tokens": 54368528}
+{"current_steps": 535, "total_steps": 2322, "loss": 0.3898594379425049, "lr": 8.750823756013498e-06, "epoch": 0.4608096468561585, "percentage": 23.04, "elapsed_time": "2:04:05", "remaining_time": "6:54:29", "throughput": 7365.67, "total_tokens": 54840552}
+{"current_steps": 540, "total_steps": 2322, "loss": 0.38840298652648925, "lr": 8.728371810397586e-06, "epoch": 0.46511627906976744, "percentage": 23.26, "elapsed_time": "2:05:11", "remaining_time": "6:53:08", "throughput": 7366.84, "total_tokens": 55337456}
+{"current_steps": 545, "total_steps": 2322, "loss": 0.3719607353210449, "lr": 8.705749243624635e-06, "epoch": 0.4694229112833764, "percentage": 23.47, "elapsed_time": "2:06:14", "remaining_time": "6:51:37", "throughput": 7373.08, "total_tokens": 55848016}
+{"current_steps": 550, "total_steps": 2322, "loss": 0.3457005262374878, "lr": 8.68295709096922e-06, "epoch": 0.4737295434969854, "percentage": 23.69, "elapsed_time": "2:07:23", "remaining_time": "6:50:27", "throughput": 7366.35, "total_tokens": 56307448}
+{"current_steps": 555, "total_steps": 2322, "loss": 0.36336798667907716, "lr": 8.659996395466648e-06, "epoch": 0.4780361757105943, "percentage": 23.9, "elapsed_time": "2:08:28", "remaining_time": "6:49:03", "throughput": 7367.1, "total_tokens": 56791296}
+{"current_steps": 560, "total_steps": 2322, "loss": 0.3849674463272095, "lr": 8.636868207865244e-06, "epoch": 0.48234280792420325, "percentage": 24.12, "elapsed_time": "2:09:36", "remaining_time": "6:47:47", "throughput": 7366.81, "total_tokens": 57286592}
+{"current_steps": 565, "total_steps": 2322, "loss": 0.3585472583770752, "lr": 8.613573586578262e-06, "epoch": 0.48664944013781225, "percentage": 24.33, "elapsed_time": "2:10:46", "remaining_time": "6:46:39", "throughput": 7363.86, "total_tokens": 57779240}
+{"current_steps": 570, "total_steps": 2322, "loss": 0.3671374797821045, "lr": 8.590113597635443e-06, "epoch": 0.4909560723514212, "percentage": 24.55, "elapsed_time": "2:11:42", "remaining_time": "6:44:51", "throughput": 7376.54, "total_tokens": 58296520}
+{"current_steps": 575, "total_steps": 2322, "loss": 0.3729721546173096, "lr": 8.56648931463423e-06, "epoch": 0.49526270456503013, "percentage": 24.76, "elapsed_time": "2:12:56", "remaining_time": "6:43:56", "throughput": 7369.72, "total_tokens": 58787768}
+{"current_steps": 580, "total_steps": 2322, "loss": 0.37489094734191897, "lr": 8.54270181869065e-06, "epoch": 0.49956933677863913, "percentage": 24.98, "elapsed_time": "2:14:10", "remaining_time": "6:42:58", "throughput": 7360.63, "total_tokens": 59253688}
+{"current_steps": 585, "total_steps": 2322, "loss": 0.3806663990020752, "lr": 8.518752198389823e-06, "epoch": 0.5038759689922481, "percentage": 25.19, "elapsed_time": "2:15:17", "remaining_time": "6:41:43", "throughput": 7363.92, "total_tokens": 59777648}
+{"current_steps": 590, "total_steps": 2322, "loss": 0.35221052169799805, "lr": 8.494641549736152e-06, "epoch": 0.508182601205857, "percentage": 25.41, "elapsed_time": "2:16:26", "remaining_time": "6:40:33", "throughput": 7361.51, "total_tokens": 60267056}
+{"current_steps": 595, "total_steps": 2322, "loss": 0.37081544399261473, "lr": 8.470370976103171e-06, "epoch": 0.512489233419466, "percentage": 25.62, "elapsed_time": "2:17:28", "remaining_time": "6:39:02", "throughput": 7365.0, "total_tokens": 60752944}
+{"current_steps": 600, "total_steps": 2322, "loss": 0.3763900279998779, "lr": 8.445941588183042e-06, "epoch": 0.5167958656330749, "percentage": 25.84, "elapsed_time": "2:18:40", "remaining_time": "6:37:59", "throughput": 7361.36, "total_tokens": 61249152}
+{"current_steps": 605, "total_steps": 2322, "loss": 0.3733165740966797, "lr": 8.421354503935733e-06, "epoch": 0.5211024978466839, "percentage": 26.06, "elapsed_time": "2:19:49", "remaining_time": "6:36:48", "throughput": 7359.28, "total_tokens": 61738896}
+{"current_steps": 610, "total_steps": 2322, "loss": 0.36464879512786863, "lr": 8.396610848537858e-06, "epoch": 0.5254091300602929, "percentage": 26.27, "elapsed_time": "2:21:08", "remaining_time": "6:36:07", "throughput": 7354.87, "total_tokens": 62285720}
+{"current_steps": 615, "total_steps": 2322, "loss": 0.3548685073852539, "lr": 8.371711754331181e-06, "epoch": 0.5297157622739018, "percentage": 26.49, "elapsed_time": "2:22:18", "remaining_time": "6:34:59", "throughput": 7351.83, "total_tokens": 62773392}
+{"current_steps": 620, "total_steps": 2322, "loss": 0.37767829895019533, "lr": 8.346658360770798e-06, "epoch": 0.5340223944875108, "percentage": 26.7, "elapsed_time": "2:23:25", "remaining_time": "6:33:42", "throughput": 7355.81, "total_tokens": 63297240}
+{"current_steps": 625, "total_steps": 2322, "loss": 0.37348289489746095, "lr": 8.321451814372998e-06, "epoch": 0.5383290267011197, "percentage": 26.92, "elapsed_time": "2:24:34", "remaining_time": "6:32:33", "throughput": 7357.39, "total_tokens": 63821872}
+{"current_steps": 630, "total_steps": 2322, "loss": 0.3854733228683472, "lr": 8.296093268662779e-06, "epoch": 0.5426356589147286, "percentage": 27.13, "elapsed_time": "2:25:43", "remaining_time": "6:31:23", "throughput": 7360.74, "total_tokens": 64361320}
+{"current_steps": 635, "total_steps": 2322, "loss": 0.3740884780883789, "lr": 8.270583884121083e-06, "epoch": 0.5469422911283376, "percentage": 27.35, "elapsed_time": "2:26:58", "remaining_time": "6:30:26", "throughput": 7357.36, "total_tokens": 64877880}
+{"current_steps": 640, "total_steps": 2322, "loss": 0.3512680768966675, "lr": 8.244924828131668e-06, "epoch": 0.5512489233419466, "percentage": 27.56, "elapsed_time": "2:28:04", "remaining_time": "6:29:08", "throughput": 7359.17, "total_tokens": 65380760}
+{"current_steps": 645, "total_steps": 2322, "loss": 0.36990017890930177, "lr": 8.219117274927696e-06, "epoch": 0.5555555555555556, "percentage": 27.78, "elapsed_time": "2:29:04", "remaining_time": "6:27:36", "throughput": 7366.25, "total_tokens": 65889880}
+{"current_steps": 650, "total_steps": 2322, "loss": 0.3620732307434082, "lr": 8.193162405537997e-06, "epoch": 0.5598621877691645, "percentage": 27.99, "elapsed_time": "2:30:13", "remaining_time": "6:26:25", "throughput": 7366.33, "total_tokens": 66397256}
+{"current_steps": 655, "total_steps": 2322, "loss": 0.37317404747009275, "lr": 8.167061407733018e-06, "epoch": 0.5641688199827735, "percentage": 28.21, "elapsed_time": "2:31:22", "remaining_time": "6:25:15", "throughput": 7366.62, "total_tokens": 66907624}
+{"current_steps": 660, "total_steps": 2322, "loss": 0.36172118186950686, "lr": 8.14081547597046e-06, "epoch": 0.5684754521963824, "percentage": 28.42, "elapsed_time": "2:32:33", "remaining_time": "6:24:11", "throughput": 7362.05, "total_tokens": 67392080}
+{"current_steps": 665, "total_steps": 2322, "loss": 0.36071798801422117, "lr": 8.114425811340635e-06, "epoch": 0.5727820844099913, "percentage": 28.64, "elapsed_time": "2:33:47", "remaining_time": "6:23:13", "throughput": 7356.12, "total_tokens": 67880624}
+{"current_steps": 670, "total_steps": 2322, "loss": 0.39921371936798095, "lr": 8.087893621511487e-06, "epoch": 0.5770887166236004, "percentage": 28.85, "elapsed_time": "2:34:56", "remaining_time": "6:22:03", "throughput": 7356.31, "total_tokens": 68390984}
+{"current_steps": 675, "total_steps": 2322, "loss": 0.38092870712280275, "lr": 8.061220120673323e-06, "epoch": 0.5813953488372093, "percentage": 29.07, "elapsed_time": "2:36:03", "remaining_time": "6:20:47", "throughput": 7357.75, "total_tokens": 68896344}
+{"current_steps": 680, "total_steps": 2322, "loss": 0.35166177749633787, "lr": 8.03440652948326e-06, "epoch": 0.5857019810508183, "percentage": 29.29, "elapsed_time": "2:37:16", "remaining_time": "6:19:46", "throughput": 7353.55, "total_tokens": 69390728}
+{"current_steps": 685, "total_steps": 2322, "loss": 0.36625821590423585, "lr": 8.007454075009352e-06, "epoch": 0.5900086132644272, "percentage": 29.5, "elapsed_time": "2:38:32", "remaining_time": "6:18:51", "throughput": 7349.98, "total_tokens": 69913800}
+{"current_steps": 690, "total_steps": 2322, "loss": 0.3945136070251465, "lr": 7.980363990674449e-06, "epoch": 0.5943152454780362, "percentage": 29.72, "elapsed_time": "2:39:45", "remaining_time": "6:17:50", "throughput": 7349.3, "total_tokens": 70444064}
+{"current_steps": 695, "total_steps": 2322, "loss": 0.36552479267120364, "lr": 7.953137516199737e-06, "epoch": 0.5986218776916451, "percentage": 29.93, "elapsed_time": "2:40:53", "remaining_time": "6:16:39", "throughput": 7348.76, "total_tokens": 70941680}
+{"current_steps": 700, "total_steps": 2322, "loss": 0.3570195198059082, "lr": 7.925775897548013e-06, "epoch": 0.602928509905254, "percentage": 30.15, "elapsed_time": "2:42:00", "remaining_time": "6:15:23", "throughput": 7349.8, "total_tokens": 71441992}
+{"current_steps": 705, "total_steps": 2322, "loss": 0.3764955043792725, "lr": 7.898280386866673e-06, "epoch": 0.6072351421188631, "percentage": 30.36, "elapsed_time": "2:43:08", "remaining_time": "6:14:10", "throughput": 7350.25, "total_tokens": 71945280}
+{"current_steps": 710, "total_steps": 2322, "loss": 0.38172245025634766, "lr": 7.87065224243039e-06, "epoch": 0.611541774332472, "percentage": 30.58, "elapsed_time": "2:44:15", "remaining_time": "6:12:56", "throughput": 7348.1, "total_tokens": 72420368}
+{"current_steps": 715, "total_steps": 2322, "loss": 0.3583367824554443, "lr": 7.842892728583557e-06, "epoch": 0.615848406546081, "percentage": 30.79, "elapsed_time": "2:45:16", "remaining_time": "6:11:28", "throughput": 7355.06, "total_tokens": 72937952}
+{"current_steps": 720, "total_steps": 2322, "loss": 0.3677781581878662, "lr": 7.815003115682405e-06, "epoch": 0.6201550387596899, "percentage": 31.01, "elapsed_time": "2:46:27", "remaining_time": "6:10:23", "throughput": 7354.21, "total_tokens": 73453368}
+{"current_steps": 725, "total_steps": 2322, "loss": 0.36302263736724855, "lr": 7.78698468003688e-06, "epoch": 0.6244616709732989, "percentage": 31.22, "elapsed_time": "2:47:34", "remaining_time": "6:09:08", "throughput": 7355.52, "total_tokens": 73959264}
+{"current_steps": 730, "total_steps": 2322, "loss": 0.37976875305175783, "lr": 7.75883870385223e-06, "epoch": 0.6287683031869078, "percentage": 31.44, "elapsed_time": "2:48:51", "remaining_time": "6:08:14", "throughput": 7348.18, "total_tokens": 74444936}
+{"current_steps": 735, "total_steps": 2322, "loss": 0.3730980157852173, "lr": 7.730566475170334e-06, "epoch": 0.6330749354005168, "percentage": 31.65, "elapsed_time": "2:49:56", "remaining_time": "6:06:56", "throughput": 7351.56, "total_tokens": 74962952}
+{"current_steps": 740, "total_steps": 2322, "loss": 0.3567670345306396, "lr": 7.702169287810751e-06, "epoch": 0.6373815676141258, "percentage": 31.87, "elapsed_time": "2:51:01", "remaining_time": "6:05:36", "throughput": 7353.11, "total_tokens": 75451216}
+{"current_steps": 745, "total_steps": 2322, "loss": 0.3866007328033447, "lr": 7.67364844131151e-06, "epoch": 0.6416881998277347, "percentage": 32.08, "elapsed_time": "2:52:02", "remaining_time": "6:04:11", "throughput": 7363.02, "total_tokens": 76008248}
+{"current_steps": 750, "total_steps": 2322, "loss": 0.3532373905181885, "lr": 7.645005240869644e-06, "epoch": 0.6459948320413437, "percentage": 32.3, "elapsed_time": "2:53:09", "remaining_time": "6:02:56", "throughput": 7365.84, "total_tokens": 76529328}
+{"current_steps": 755, "total_steps": 2322, "loss": 0.3541311502456665, "lr": 7.616240997281465e-06, "epoch": 0.6503014642549526, "percentage": 32.52, "elapsed_time": "2:54:26", "remaining_time": "6:02:03", "throughput": 7361.22, "total_tokens": 77047776}
+{"current_steps": 760, "total_steps": 2322, "loss": 0.36530122756958006, "lr": 7.587357026882563e-06, "epoch": 0.6546080964685616, "percentage": 32.73, "elapsed_time": "2:55:41", "remaining_time": "6:01:05", "throughput": 7359.9, "total_tokens": 77585376}
+{"current_steps": 765, "total_steps": 2322, "loss": 0.3949885845184326, "lr": 7.558354651487583e-06, "epoch": 0.6589147286821705, "percentage": 32.95, "elapsed_time": "2:56:51", "remaining_time": "5:59:56", "throughput": 7359.76, "total_tokens": 78094440}
+{"current_steps": 770, "total_steps": 2322, "loss": 0.3596529006958008, "lr": 7.529235198329723e-06, "epoch": 0.6632213608957795, "percentage": 33.16, "elapsed_time": "2:58:11", "remaining_time": "5:59:09", "throughput": 7354.92, "total_tokens": 78635856}
+{"current_steps": 775, "total_steps": 2322, "loss": 0.35566143989562987, "lr": 7.500000000000001e-06, "epoch": 0.6675279931093885, "percentage": 33.38, "elapsed_time": "2:59:16", "remaining_time": "5:57:51", "throughput": 7357.08, "total_tokens": 79138600}
+{"current_steps": 780, "total_steps": 2322, "loss": 0.36934409141540525, "lr": 7.470650394386274e-06, "epoch": 0.6718346253229974, "percentage": 33.59, "elapsed_time": "3:00:25", "remaining_time": "5:56:40", "throughput": 7357.01, "total_tokens": 79642424}
+{"current_steps": 785, "total_steps": 2322, "loss": 0.37495574951171873, "lr": 7.441187724612007e-06, "epoch": 0.6761412575366064, "percentage": 33.81, "elapsed_time": "3:01:30", "remaining_time": "5:55:23", "throughput": 7359.8, "total_tokens": 80152744}
+{"current_steps": 790, "total_steps": 2322, "loss": 0.37909436225891113, "lr": 7.4116133389748115e-06, "epoch": 0.6804478897502153, "percentage": 34.02, "elapsed_time": "3:02:24", "remaining_time": "5:53:43", "throughput": 7373.27, "total_tokens": 80693400}
+{"current_steps": 795, "total_steps": 2322, "loss": 0.361489462852478, "lr": 7.381928590884741e-06, "epoch": 0.6847545219638242, "percentage": 34.24, "elapsed_time": "3:03:29", "remaining_time": "5:52:27", "throughput": 7373.44, "total_tokens": 81180032}
+{"current_steps": 800, "total_steps": 2322, "loss": 0.34208495616912843, "lr": 7.35213483880236e-06, "epoch": 0.6890611541774333, "percentage": 34.45, "elapsed_time": "3:04:42", "remaining_time": "5:51:25", "throughput": 7373.18, "total_tokens": 81715960}
+{"current_steps": 805, "total_steps": 2322, "loss": 0.3672114133834839, "lr": 7.322233446176571e-06, "epoch": 0.6933677863910422, "percentage": 34.67, "elapsed_time": "3:05:53", "remaining_time": "5:50:18", "throughput": 7369.41, "total_tokens": 82195808}
+{"current_steps": 810, "total_steps": 2322, "loss": 0.3744704008102417, "lr": 7.29222578138222e-06, "epoch": 0.6976744186046512, "percentage": 34.88, "elapsed_time": "3:07:01", "remaining_time": "5:49:06", "throughput": 7367.98, "total_tokens": 82680488}
+{"current_steps": 815, "total_steps": 2322, "loss": 0.373249888420105, "lr": 7.2621132176574774e-06, "epoch": 0.7019810508182601, "percentage": 35.1, "elapsed_time": "3:08:12", "remaining_time": "5:47:59", "throughput": 7368.61, "total_tokens": 83207016}
+{"current_steps": 820, "total_steps": 2322, "loss": 0.34656834602355957, "lr": 7.231897133040997e-06, "epoch": 0.7062876830318691, "percentage": 35.31, "elapsed_time": "3:09:26", "remaining_time": "5:46:59", "throughput": 7364.53, "total_tokens": 83705464}
+{"current_steps": 825, "total_steps": 2322, "loss": 0.36565725803375243, "lr": 7.201578910308848e-06, "epoch": 0.710594315245478, "percentage": 35.53, "elapsed_time": "3:10:39", "remaining_time": "5:45:56", "throughput": 7362.84, "total_tokens": 84223624}
+{"current_steps": 830, "total_steps": 2322, "loss": 0.3640167474746704, "lr": 7.1711599369112385e-06, "epoch": 0.7149009474590869, "percentage": 35.75, "elapsed_time": "3:11:38", "remaining_time": "5:44:29", "throughput": 7369.19, "total_tokens": 84733008}
+{"current_steps": 835, "total_steps": 2322, "loss": 0.34748039245605467, "lr": 7.14064160490902e-06, "epoch": 0.719207579672696, "percentage": 35.96, "elapsed_time": "3:12:50", "remaining_time": "5:43:24", "throughput": 7367.69, "total_tokens": 85244448}
+{"current_steps": 840, "total_steps": 2322, "loss": 0.353003716468811, "lr": 7.110025310909981e-06, "epoch": 0.7235142118863049, "percentage": 36.18, "elapsed_time": "3:13:58", "remaining_time": "5:42:14", "throughput": 7368.71, "total_tokens": 85762344}
+{"current_steps": 845, "total_steps": 2322, "loss": 0.37342329025268556, "lr": 7.079312456004941e-06, "epoch": 0.7278208440999139, "percentage": 36.39, "elapsed_time": "3:15:16", "remaining_time": "5:41:18", "throughput": 7364.98, "total_tokens": 86289224}
+{"current_steps": 850, "total_steps": 2322, "loss": 0.3299787759780884, "lr": 7.048504445703623e-06, "epoch": 0.7321274763135228, "percentage": 36.61, "elapsed_time": "3:16:25", "remaining_time": "5:40:09", "throughput": 7364.74, "total_tokens": 86796928}
+{"current_steps": 855, "total_steps": 2322, "loss": 0.36838767528533933, "lr": 7.017602689870345e-06, "epoch": 0.7364341085271318, "percentage": 36.82, "elapsed_time": "3:17:30", "remaining_time": "5:38:53", "throughput": 7368.24, "total_tokens": 87320496}
+{"current_steps": 860, "total_steps": 2322, "loss": 0.3541435718536377, "lr": 6.986608602659486e-06, "epoch": 0.7407407407407407, "percentage": 37.04, "elapsed_time": "3:18:37", "remaining_time": "5:37:39", "throughput": 7374.08, "total_tokens": 87879104}
+{"current_steps": 865, "total_steps": 2322, "loss": 0.35468919277191163, "lr": 6.95552360245078e-06, "epoch": 0.7450473729543498, "percentage": 37.25, "elapsed_time": "3:19:46", "remaining_time": "5:36:29", "throughput": 7373.04, "total_tokens": 88373816}
+{"current_steps": 870, "total_steps": 2322, "loss": 0.36195812225341795, "lr": 6.92434911178441e-06, "epoch": 0.7493540051679587, "percentage": 37.47, "elapsed_time": "3:20:56", "remaining_time": "5:35:21", "throughput": 7375.3, "total_tokens": 88916672}
+{"current_steps": 875, "total_steps": 2322, "loss": 0.3709383726119995, "lr": 6.893086557295896e-06, "epoch": 0.7536606373815676, "percentage": 37.68, "elapsed_time": "3:22:00", "remaining_time": "5:34:03", "throughput": 7378.25, "total_tokens": 89428224}
+{"current_steps": 880, "total_steps": 2322, "loss": 0.362162446975708, "lr": 6.861737369650818e-06, "epoch": 0.7579672695951766, "percentage": 37.9, "elapsed_time": "3:23:12", "remaining_time": "5:32:58", "throughput": 7375.53, "total_tokens": 89924488}
+{"current_steps": 885, "total_steps": 2322, "loss": 0.36125888824462893, "lr": 6.830302983479344e-06, "epoch": 0.7622739018087855, "percentage": 38.11, "elapsed_time": "3:24:20", "remaining_time": "5:31:47", "throughput": 7379.01, "total_tokens": 90470424}
+{"current_steps": 890, "total_steps": 2322, "loss": 0.35643205642700193, "lr": 6.7987848373105705e-06, "epoch": 0.7665805340223945, "percentage": 38.33, "elapsed_time": "3:25:36", "remaining_time": "5:30:49", "throughput": 7376.0, "total_tokens": 90996680}
+{"current_steps": 895, "total_steps": 2322, "loss": 0.34829187393188477, "lr": 6.767184373506698e-06, "epoch": 0.7708871662360034, "percentage": 38.54, "elapsed_time": "3:26:27", "remaining_time": "5:29:10", "throughput": 7384.91, "total_tokens": 91480544}
+{"current_steps": 900, "total_steps": 2322, "loss": 0.3527660846710205, "lr": 6.73550303819702e-06, "epoch": 0.7751937984496124, "percentage": 38.76, "elapsed_time": "3:27:38", "remaining_time": "5:28:04", "throughput": 7382.02, "total_tokens": 91969264}
+{"current_steps": 905, "total_steps": 2322, "loss": 0.35948958396911623, "lr": 6.70374228121175e-06, "epoch": 0.7795004306632214, "percentage": 38.98, "elapsed_time": "3:28:49", "remaining_time": "5:26:58", "throughput": 7379.09, "total_tokens": 92456304}
+{"current_steps": 910, "total_steps": 2322, "loss": 0.35416080951690676, "lr": 6.671903556015664e-06, "epoch": 0.7838070628768303, "percentage": 39.19, "elapsed_time": "3:30:02", "remaining_time": "5:25:54", "throughput": 7378.37, "total_tokens": 92982984}
+{"current_steps": 915, "total_steps": 2322, "loss": 0.34634552001953123, "lr": 6.639988319641592e-06, "epoch": 0.7881136950904393, "percentage": 39.41, "elapsed_time": "3:31:11", "remaining_time": "5:24:44", "throughput": 7376.01, "total_tokens": 93461984}
+{"current_steps": 920, "total_steps": 2322, "loss": 0.37337236404418944, "lr": 6.607998032623741e-06, "epoch": 0.7924203273040482, "percentage": 39.62, "elapsed_time": "3:32:15", "remaining_time": "5:23:27", "throughput": 7379.64, "total_tokens": 93979864}
+{"current_steps": 925, "total_steps": 2322, "loss": 0.3476158380508423, "lr": 6.57593415893085e-06, "epoch": 0.7967269595176572, "percentage": 39.84, "elapsed_time": "3:33:21", "remaining_time": "5:22:14", "throughput": 7378.85, "total_tokens": 94462232}
+{"current_steps": 930, "total_steps": 2322, "loss": 0.35996217727661134, "lr": 6.5437981658992e-06, "epoch": 0.8010335917312662, "percentage": 40.05, "elapsed_time": "3:34:25", "remaining_time": "5:20:56", "throughput": 7383.76, "total_tokens": 94992816}
+{"current_steps": 935, "total_steps": 2322, "loss": 0.36520819664001464, "lr": 6.511591524165465e-06, "epoch": 0.8053402239448751, "percentage": 40.27, "elapsed_time": "3:35:37", "remaining_time": "5:19:51", "throughput": 7380.8, "total_tokens": 95488328}
+{"current_steps": 940, "total_steps": 2322, "loss": 0.36473860740661623, "lr": 6.479315707599407e-06, "epoch": 0.8096468561584841, "percentage": 40.48, "elapsed_time": "3:36:36", "remaining_time": "5:18:27", "throughput": 7383.15, "total_tokens": 95954480}
+{"current_steps": 945, "total_steps": 2322, "loss": 0.3522223472595215, "lr": 6.446972193236433e-06, "epoch": 0.813953488372093, "percentage": 40.7, "elapsed_time": "3:37:36", "remaining_time": "5:17:04", "throughput": 7390.47, "total_tokens": 96490896}
+{"current_steps": 950, "total_steps": 2322, "loss": 0.3675665855407715, "lr": 6.414562461209994e-06, "epoch": 0.818260120585702, "percentage": 40.91, "elapsed_time": "3:38:46", "remaining_time": "5:15:57", "throughput": 7392.02, "total_tokens": 97032496}
+{"current_steps": 955, "total_steps": 2322, "loss": 0.3679534435272217, "lr": 6.3820879946838585e-06, "epoch": 0.8225667527993109, "percentage": 41.13, "elapsed_time": "3:40:01", "remaining_time": "5:14:57", "throughput": 7388.52, "total_tokens": 97543080}
+{"current_steps": 960, "total_steps": 2322, "loss": 0.3499560594558716, "lr": 6.349550279784231e-06, "epoch": 0.8268733850129198, "percentage": 41.34, "elapsed_time": "3:41:14", "remaining_time": "5:13:52", "throughput": 7384.36, "total_tokens": 98020152}
+{"current_steps": 965, "total_steps": 2322, "loss": 0.3565502166748047, "lr": 6.316950805531746e-06, "epoch": 0.8311800172265289, "percentage": 41.56, "elapsed_time": "3:42:23", "remaining_time": "5:12:43", "throughput": 7381.76, "total_tokens": 98498816}
+{"current_steps": 970, "total_steps": 2322, "loss": 0.36150004863739016, "lr": 6.284291063773331e-06, "epoch": 0.8354866494401378, "percentage": 41.77, "elapsed_time": "3:43:39", "remaining_time": "5:11:44", "throughput": 7380.75, "total_tokens": 99045784}
+{"current_steps": 975, "total_steps": 2322, "loss": 0.36032235622406006, "lr": 6.251572549113925e-06, "epoch": 0.8397932816537468, "percentage": 41.99, "elapsed_time": "3:44:47", "remaining_time": "5:10:33", "throughput": 7383.26, "total_tokens": 99579432}
+{"current_steps": 980, "total_steps": 2322, "loss": 0.3623358726501465, "lr": 6.218796758848085e-06, "epoch": 0.8440999138673557, "percentage": 42.2, "elapsed_time": "3:45:56", "remaining_time": "5:09:24", "throughput": 7383.15, "total_tokens": 100089992}
+{"current_steps": 985, "total_steps": 2322, "loss": 0.36282594203948976, "lr": 6.185965192891472e-06, "epoch": 0.8484065460809647, "percentage": 42.42, "elapsed_time": "3:46:54", "remaining_time": "5:07:59", "throughput": 7390.56, "total_tokens": 100618600}
+{"current_steps": 990, "total_steps": 2322, "loss": 0.3642581939697266, "lr": 6.153079353712201e-06, "epoch": 0.8527131782945736, "percentage": 42.64, "elapsed_time": "3:47:57", "remaining_time": "5:06:42", "throughput": 7392.59, "total_tokens": 101111776}
+{"current_steps": 995, "total_steps": 2322, "loss": 0.3563115119934082, "lr": 6.120140746262091e-06, "epoch": 0.8570198105081827, "percentage": 42.85, "elapsed_time": "3:49:10", "remaining_time": "5:05:38", "throughput": 7389.29, "total_tokens": 101603528}
+{"current_steps": 1000, "total_steps": 2322, "loss": 0.3616267442703247, "lr": 6.087150877907786e-06, "epoch": 0.8613264427217916, "percentage": 43.07, "elapsed_time": "3:50:17", "remaining_time": "5:04:26", "throughput": 7392.74, "total_tokens": 102147912}
+{"current_steps": 1005, "total_steps": 2322, "loss": 0.3602945804595947, "lr": 6.054111258361782e-06, "epoch": 0.8656330749354005, "percentage": 43.28, "elapsed_time": "3:51:20", "remaining_time": "5:03:09", "throughput": 7395.93, "total_tokens": 102659472}
+{"current_steps": 1010, "total_steps": 2322, "loss": 0.34914188385009765, "lr": 6.021023399613335e-06, "epoch": 0.8699397071490095, "percentage": 43.5, "elapsed_time": "3:52:31", "remaining_time": "5:02:03", "throughput": 7395.42, "total_tokens": 103179440}
+{"current_steps": 1015, "total_steps": 2322, "loss": 0.3555158138275146, "lr": 5.987888815859266e-06, "epoch": 0.8742463393626184, "percentage": 43.71, "elapsed_time": "3:53:42", "remaining_time": "5:00:56", "throughput": 7394.58, "total_tokens": 103690584}
+{"current_steps": 1020, "total_steps": 2322, "loss": 0.33635711669921875, "lr": 5.954709023434669e-06, "epoch": 0.8785529715762274, "percentage": 43.93, "elapsed_time": "3:54:45", "remaining_time": "4:59:39", "throughput": 7396.27, "total_tokens": 104178112}
+{"current_steps": 1025, "total_steps": 2322, "loss": 0.35177910327911377, "lr": 5.921485540743516e-06, "epoch": 0.8828596037898363, "percentage": 44.14, "elapsed_time": "3:56:00", "remaining_time": "4:58:37", "throughput": 7393.97, "total_tokens": 104699576}
+{"current_steps": 1030, "total_steps": 2322, "loss": 0.36098456382751465, "lr": 5.888219888189176e-06, "epoch": 0.8871662360034454, "percentage": 44.36, "elapsed_time": "3:56:53", "remaining_time": "4:57:08", "throughput": 7402.65, "total_tokens": 105214336}
+{"current_steps": 1035, "total_steps": 2322, "loss": 0.37163600921630857, "lr": 5.854913588104832e-06, "epoch": 0.8914728682170543, "percentage": 44.57, "elapsed_time": "3:57:53", "remaining_time": "4:55:48", "throughput": 7406.3, "total_tokens": 105712752}
+{"current_steps": 1040, "total_steps": 2322, "loss": 0.37590489387512205, "lr": 5.82156816468382e-06, "epoch": 0.8957795004306632, "percentage": 44.79, "elapsed_time": "3:59:01", "remaining_time": "4:54:38", "throughput": 7406.05, "total_tokens": 106215336}
+{"current_steps": 1045, "total_steps": 2322, "loss": 0.35026397705078127, "lr": 5.788185143909868e-06, "epoch": 0.9000861326442722, "percentage": 45.0, "elapsed_time": "4:00:14", "remaining_time": "4:53:34", "throughput": 7404.77, "total_tokens": 106737504}
+{"current_steps": 1050, "total_steps": 2322, "loss": 0.3652517080307007, "lr": 5.754766053487274e-06, "epoch": 0.9043927648578811, "percentage": 45.22, "elapsed_time": "4:01:24", "remaining_time": "4:52:27", "throughput": 7405.04, "total_tokens": 107261064}
+{"current_steps": 1055, "total_steps": 2322, "loss": 0.3495129108428955, "lr": 5.721312422770984e-06, "epoch": 0.9086993970714901, "percentage": 45.43, "elapsed_time": "4:02:44", "remaining_time": "4:51:31", "throughput": 7400.66, "total_tokens": 107787160}
+{"current_steps": 1060, "total_steps": 2322, "loss": 0.34899396896362306, "lr": 5.68782578269661e-06, "epoch": 0.9130060292850991, "percentage": 45.65, "elapsed_time": "4:03:57", "remaining_time": "4:50:27", "throughput": 7397.11, "total_tokens": 108278376}
+{"current_steps": 1065, "total_steps": 2322, "loss": 0.35709681510925295, "lr": 5.6543076657103705e-06, "epoch": 0.917312661498708, "percentage": 45.87, "elapsed_time": "4:05:03", "remaining_time": "4:49:14", "throughput": 7395.39, "total_tokens": 108741080}
+{"current_steps": 1070, "total_steps": 2322, "loss": 0.35440216064453123, "lr": 5.620759605698953e-06, "epoch": 0.921619293712317, "percentage": 46.08, "elapsed_time": "4:06:23", "remaining_time": "4:48:18", "throughput": 7390.05, "total_tokens": 109252016}
+{"current_steps": 1075, "total_steps": 2322, "loss": 0.3724498748779297, "lr": 5.587183137919332e-06, "epoch": 0.9259259259259259, "percentage": 46.3, "elapsed_time": "4:07:38", "remaining_time": "4:47:15", "throughput": 7387.45, "total_tokens": 109763232}
+{"current_steps": 1080, "total_steps": 2322, "loss": 0.35597658157348633, "lr": 5.553579798928503e-06, "epoch": 0.9302325581395349, "percentage": 46.51, "elapsed_time": "4:08:40", "remaining_time": "4:45:58", "throughput": 7389.81, "total_tokens": 110258264}
+{"current_steps": 1085, "total_steps": 2322, "loss": 0.366562032699585, "lr": 5.519951126513164e-06, "epoch": 0.9345391903531438, "percentage": 46.73, "elapsed_time": "4:09:49", "remaining_time": "4:44:49", "throughput": 7389.22, "total_tokens": 110761536}
+{"current_steps": 1090, "total_steps": 2322, "loss": 0.3656296730041504, "lr": 5.486298659619346e-06, "epoch": 0.9388458225667528, "percentage": 46.94, "elapsed_time": "4:10:50", "remaining_time": "4:43:30", "throughput": 7390.44, "total_tokens": 111228400}
+{"current_steps": 1095, "total_steps": 2322, "loss": 0.3466034412384033, "lr": 5.452623938281983e-06, "epoch": 0.9431524547803618, "percentage": 47.16, "elapsed_time": "4:12:05", "remaining_time": "4:42:28", "throughput": 7387.29, "total_tokens": 111733504}
+{"current_steps": 1100, "total_steps": 2322, "loss": 0.36214759349823, "lr": 5.418928503554436e-06, "epoch": 0.9474590869939707, "percentage": 47.37, "elapsed_time": "4:13:13", "remaining_time": "4:41:18", "throughput": 7386.41, "total_tokens": 112225696}
+{"current_steps": 1105, "total_steps": 2322, "loss": 0.3605997085571289, "lr": 5.385213897437975e-06, "epoch": 0.9517657192075797, "percentage": 47.59, "elapsed_time": "4:14:12", "remaining_time": "4:39:58", "throughput": 7392.02, "total_tokens": 112749656}
+{"current_steps": 1110, "total_steps": 2322, "loss": 0.3977372407913208, "lr": 5.351481662811208e-06, "epoch": 0.9560723514211886, "percentage": 47.8, "elapsed_time": "4:15:15", "remaining_time": "4:38:43", "throughput": 7398.08, "total_tokens": 113307680}
+{"current_steps": 1115, "total_steps": 2322, "loss": 0.3589299201965332, "lr": 5.3177333433594734e-06, "epoch": 0.9603789836347976, "percentage": 48.02, "elapsed_time": "4:16:26", "remaining_time": "4:37:35", "throughput": 7397.32, "total_tokens": 113816672}
+{"current_steps": 1120, "total_steps": 2322, "loss": 0.35573978424072267, "lr": 5.283970483504198e-06, "epoch": 0.9646856158484065, "percentage": 48.23, "elapsed_time": "4:17:40", "remaining_time": "4:36:32", "throughput": 7394.89, "total_tokens": 114328160}
+{"current_steps": 1125, "total_steps": 2322, "loss": 0.3410700798034668, "lr": 5.2501946283322204e-06, "epoch": 0.9689922480620154, "percentage": 48.45, "elapsed_time": "4:18:42", "remaining_time": "4:35:15", "throughput": 7395.18, "total_tokens": 114790904}
+{"current_steps": 1130, "total_steps": 2322, "loss": 0.3575610876083374, "lr": 5.216407323525083e-06, "epoch": 0.9732988802756245, "percentage": 48.66, "elapsed_time": "4:19:49", "remaining_time": "4:34:04", "throughput": 7395.1, "total_tokens": 115286400}
+{"current_steps": 1135, "total_steps": 2322, "loss": 0.35471277236938475, "lr": 5.182610115288296e-06, "epoch": 0.9776055124892334, "percentage": 48.88, "elapsed_time": "4:21:04", "remaining_time": "4:33:01", "throughput": 7393.91, "total_tokens": 115819512}
+{"current_steps": 1140, "total_steps": 2322, "loss": 0.35717573165893557, "lr": 5.148804550280578e-06, "epoch": 0.9819121447028424, "percentage": 49.1, "elapsed_time": "4:22:07", "remaining_time": "4:31:46", "throughput": 7395.1, "total_tokens": 116305168}
+{"current_steps": 1145, "total_steps": 2322, "loss": 0.346773099899292, "lr": 5.114992175543084e-06, "epoch": 0.9862187769164513, "percentage": 49.31, "elapsed_time": "4:23:09", "remaining_time": "4:30:30", "throughput": 7394.51, "total_tokens": 116753064}
+{"current_steps": 1150, "total_steps": 2322, "loss": 0.36169114112854006, "lr": 5.081174538428596e-06, "epoch": 0.9905254091300603, "percentage": 49.53, "elapsed_time": "4:24:19", "remaining_time": "4:29:23", "throughput": 7394.38, "total_tokens": 117272904}
+{"current_steps": 1155, "total_steps": 2322, "loss": 0.34629039764404296, "lr": 5.047353186530718e-06, "epoch": 0.9948320413436692, "percentage": 49.74, "elapsed_time": "4:25:25", "remaining_time": "4:28:10", "throughput": 7394.43, "total_tokens": 117758136}
+{"current_steps": 1160, "total_steps": 2322, "loss": 0.35693745613098143, "lr": 5.0135296676130576e-06, "epoch": 0.9991386735572783, "percentage": 49.96, "elapsed_time": "4:26:37", "remaining_time": "4:27:05", "throughput": 7393.29, "total_tokens": 118275664}
+{"current_steps": 1165, "total_steps": 2322, "loss": 0.3211682796478271, "lr": 4.979705529538385e-06, "epoch": 1.0034453057708872, "percentage": 50.17, "elapsed_time": "4:27:47", "remaining_time": "4:25:56", "throughput": 7392.18, "total_tokens": 118772088}
+{"current_steps": 1170, "total_steps": 2322, "loss": 0.3132964611053467, "lr": 4.9458823201978124e-06, "epoch": 1.0077519379844961, "percentage": 50.39, "elapsed_time": "4:29:06", "remaining_time": "4:24:57", "throughput": 7390.32, "total_tokens": 119324520}
+{"current_steps": 1175, "total_steps": 2322, "loss": 0.32307000160217286, "lr": 4.912061587439944e-06, "epoch": 1.012058570198105, "percentage": 50.6, "elapsed_time": "4:30:20", "remaining_time": "4:23:54", "throughput": 7389.45, "total_tokens": 119862352}
+{"current_steps": 1180, "total_steps": 2322, "loss": 0.3019747257232666, "lr": 4.8782448790000525e-06, "epoch": 1.016365202411714, "percentage": 50.82, "elapsed_time": "4:31:28", "remaining_time": "4:22:44", "throughput": 7388.79, "total_tokens": 120353472}
+{"current_steps": 1185, "total_steps": 2322, "loss": 0.3043541669845581, "lr": 4.8444337424292445e-06, "epoch": 1.020671834625323, "percentage": 51.03, "elapsed_time": "4:32:49", "remaining_time": "4:21:46", "throughput": 7383.42, "total_tokens": 120865584}
+{"current_steps": 1190, "total_steps": 2322, "loss": 0.32289729118347166, "lr": 4.81062972502364e-06, "epoch": 1.024978466838932, "percentage": 51.25, "elapsed_time": "4:34:05", "remaining_time": "4:20:43", "throughput": 7380.98, "total_tokens": 121382112}
+{"current_steps": 1195, "total_steps": 2322, "loss": 0.3106250524520874, "lr": 4.7768343737535694e-06, "epoch": 1.0292850990525408, "percentage": 51.46, "elapsed_time": "4:35:20", "remaining_time": "4:19:40", "throughput": 7381.35, "total_tokens": 121946272}
+{"current_steps": 1200, "total_steps": 2322, "loss": 0.30351016521453855, "lr": 4.743049235192771e-06, "epoch": 1.0335917312661498, "percentage": 51.68, "elapsed_time": "4:36:32", "remaining_time": "4:18:33", "throughput": 7380.22, "total_tokens": 122456216}
+{"current_steps": 1205, "total_steps": 2322, "loss": 0.30516183376312256, "lr": 4.7092758554476215e-06, "epoch": 1.0378983634797587, "percentage": 51.89, "elapsed_time": "4:37:42", "remaining_time": "4:17:26", "throughput": 7381.5, "total_tokens": 122997032}
+{"current_steps": 1210, "total_steps": 2322, "loss": 0.30727853775024416, "lr": 4.6755157800863826e-06, "epoch": 1.0422049956933679, "percentage": 52.11, "elapsed_time": "4:38:53", "remaining_time": "4:16:18", "throughput": 7380.64, "total_tokens": 123504248}
+{"current_steps": 1215, "total_steps": 2322, "loss": 0.2971274137496948, "lr": 4.641770554068465e-06, "epoch": 1.0465116279069768, "percentage": 52.33, "elapsed_time": "4:39:58", "remaining_time": "4:15:05", "throughput": 7382.8, "total_tokens": 124021952}
+{"current_steps": 1220, "total_steps": 2322, "loss": 0.31333255767822266, "lr": 4.608041721673731e-06, "epoch": 1.0508182601205858, "percentage": 52.54, "elapsed_time": "4:41:06", "remaining_time": "4:13:54", "throughput": 7383.99, "total_tokens": 124538536}
+{"current_steps": 1225, "total_steps": 2322, "loss": 0.33772854804992675, "lr": 4.574330826431822e-06, "epoch": 1.0551248923341947, "percentage": 52.76, "elapsed_time": "4:42:17", "remaining_time": "4:12:47", "throughput": 7382.03, "total_tokens": 125033152}
+{"current_steps": 1230, "total_steps": 2322, "loss": 0.3157934665679932, "lr": 4.540639411051528e-06, "epoch": 1.0594315245478036, "percentage": 52.97, "elapsed_time": "4:43:20", "remaining_time": "4:11:32", "throughput": 7386.94, "total_tokens": 125580048}
+{"current_steps": 1235, "total_steps": 2322, "loss": 0.304073166847229, "lr": 4.506969017350178e-06, "epoch": 1.0637381567614126, "percentage": 53.19, "elapsed_time": "4:44:28", "remaining_time": "4:10:22", "throughput": 7386.86, "total_tokens": 126082048}
+{"current_steps": 1240, "total_steps": 2322, "loss": 0.29829227924346924, "lr": 4.473321186183091e-06, "epoch": 1.0680447889750215, "percentage": 53.4, "elapsed_time": "4:45:33", "remaining_time": "4:09:10", "throughput": 7391.0, "total_tokens": 126635088}
+{"current_steps": 1245, "total_steps": 2322, "loss": 0.2966867685317993, "lr": 4.439697457373055e-06, "epoch": 1.0723514211886305, "percentage": 53.62, "elapsed_time": "4:46:32", "remaining_time": "4:07:52", "throughput": 7393.47, "total_tokens": 127113288}
+{"current_steps": 1250, "total_steps": 2322, "loss": 0.30222883224487307, "lr": 4.40609936963987e-06, "epoch": 1.0766580534022394, "percentage": 53.83, "elapsed_time": "4:47:28", "remaining_time": "4:06:32", "throughput": 7399.93, "total_tokens": 127635064}
+{"current_steps": 1255, "total_steps": 2322, "loss": 0.3315183877944946, "lr": 4.37252846052992e-06, "epoch": 1.0809646856158484, "percentage": 54.05, "elapsed_time": "4:48:33", "remaining_time": "4:05:19", "throughput": 7403.97, "total_tokens": 128185744}
+{"current_steps": 1260, "total_steps": 2322, "loss": 0.31603388786315917, "lr": 4.338986266345819e-06, "epoch": 1.0852713178294573, "percentage": 54.26, "elapsed_time": "4:49:34", "remaining_time": "4:04:04", "throughput": 7407.0, "total_tokens": 128696104}
+{"current_steps": 1265, "total_steps": 2322, "loss": 0.29094948768615725, "lr": 4.305474322076102e-06, "epoch": 1.0895779500430662, "percentage": 54.48, "elapsed_time": "4:50:43", "remaining_time": "4:02:55", "throughput": 7406.08, "total_tokens": 129191096}
+{"current_steps": 1270, "total_steps": 2322, "loss": 0.3133362293243408, "lr": 4.271994161324977e-06, "epoch": 1.0938845822566754, "percentage": 54.69, "elapsed_time": "4:51:52", "remaining_time": "4:01:46", "throughput": 7404.55, "total_tokens": 129671408}
+{"current_steps": 1275, "total_steps": 2322, "loss": 0.3118987321853638, "lr": 4.238547316242149e-06, "epoch": 1.0981912144702843, "percentage": 54.91, "elapsed_time": "4:53:04", "remaining_time": "4:00:39", "throughput": 7402.08, "total_tokens": 130159968}
+{"current_steps": 1280, "total_steps": 2322, "loss": 0.31302499771118164, "lr": 4.205135317452701e-06, "epoch": 1.1024978466838933, "percentage": 55.12, "elapsed_time": "4:54:14", "remaining_time": "3:59:32", "throughput": 7402.9, "total_tokens": 130697176}
+{"current_steps": 1285, "total_steps": 2322, "loss": 0.3269864797592163, "lr": 4.171759693987046e-06, "epoch": 1.1068044788975022, "percentage": 55.34, "elapsed_time": "4:55:14", "remaining_time": "3:58:15", "throughput": 7405.78, "total_tokens": 131192368}
+{"current_steps": 1290, "total_steps": 2322, "loss": 0.3242049217224121, "lr": 4.1384219732109544e-06, "epoch": 1.1111111111111112, "percentage": 55.56, "elapsed_time": "4:56:22", "remaining_time": "3:57:06", "throughput": 7406.78, "total_tokens": 131711328}
+{"current_steps": 1295, "total_steps": 2322, "loss": 0.30824267864227295, "lr": 4.105123680755667e-06, "epoch": 1.11541774332472, "percentage": 55.77, "elapsed_time": "4:57:39", "remaining_time": "3:56:03", "throughput": 7404.94, "total_tokens": 132247936}
+{"current_steps": 1300, "total_steps": 2322, "loss": 0.3098948001861572, "lr": 4.071866340448062e-06, "epoch": 1.119724375538329, "percentage": 55.99, "elapsed_time": "4:58:41", "remaining_time": "3:54:49", "throughput": 7408.48, "total_tokens": 132771952}
+{"current_steps": 1305, "total_steps": 2322, "loss": 0.312363338470459, "lr": 4.03865147424093e-06, "epoch": 1.124031007751938, "percentage": 56.2, "elapsed_time": "5:00:00", "remaining_time": "3:53:47", "throughput": 7406.53, "total_tokens": 133320696}
+{"current_steps": 1310, "total_steps": 2322, "loss": 0.3138570785522461, "lr": 4.0054806021433245e-06, "epoch": 1.128337639965547, "percentage": 56.42, "elapsed_time": "5:01:03", "remaining_time": "3:52:34", "throughput": 7407.99, "total_tokens": 133811104}
+{"current_steps": 1315, "total_steps": 2322, "loss": 0.2930002212524414, "lr": 3.9723552421509975e-06, "epoch": 1.1326442721791559, "percentage": 56.63, "elapsed_time": "5:02:13", "remaining_time": "3:51:25", "throughput": 7407.46, "total_tokens": 134319512}
+{"current_steps": 1320, "total_steps": 2322, "loss": 0.29963154792785646, "lr": 3.939276910176939e-06, "epoch": 1.1369509043927648, "percentage": 56.85, "elapsed_time": "5:03:33", "remaining_time": "3:50:25", "throughput": 7404.24, "total_tokens": 134854032}
+{"current_steps": 1325, "total_steps": 2322, "loss": 0.31452503204345705, "lr": 3.906247119981995e-06, "epoch": 1.1412575366063737, "percentage": 57.06, "elapsed_time": "5:04:32", "remaining_time": "3:49:08", "throughput": 7408.55, "total_tokens": 135369192}
+{"current_steps": 1330, "total_steps": 2322, "loss": 0.3095171213150024, "lr": 3.873267383105604e-06, "epoch": 1.1455641688199827, "percentage": 57.28, "elapsed_time": "5:05:36", "remaining_time": "3:47:56", "throughput": 7412.2, "total_tokens": 135915648}
+{"current_steps": 1335, "total_steps": 2322, "loss": 0.32448587417602537, "lr": 3.840339208796611e-06, "epoch": 1.1498708010335918, "percentage": 57.49, "elapsed_time": "5:06:50", "remaining_time": "3:46:51", "throughput": 7411.08, "total_tokens": 136438696}
+{"current_steps": 1340, "total_steps": 2322, "loss": 0.3084967613220215, "lr": 3.8074641039442138e-06, "epoch": 1.1541774332472006, "percentage": 57.71, "elapsed_time": "5:07:58", "remaining_time": "3:45:41", "throughput": 7411.77, "total_tokens": 136955600}
+{"current_steps": 1345, "total_steps": 2322, "loss": 0.28108384609222414, "lr": 3.774643573008995e-06, "epoch": 1.1584840654608097, "percentage": 57.92, "elapsed_time": "5:09:06", "remaining_time": "3:44:32", "throughput": 7411.18, "total_tokens": 137454648}
+{"current_steps": 1350, "total_steps": 2322, "loss": 0.3180356502532959, "lr": 3.741879117954077e-06, "epoch": 1.1627906976744187, "percentage": 58.14, "elapsed_time": "5:10:16", "remaining_time": "3:43:23", "throughput": 7411.48, "total_tokens": 137974504}
+{"current_steps": 1355, "total_steps": 2322, "loss": 0.2953655481338501, "lr": 3.709172238176384e-06, "epoch": 1.1670973298880276, "percentage": 58.35, "elapsed_time": "5:11:20", "remaining_time": "3:42:11", "throughput": 7412.88, "total_tokens": 138476768}
+{"current_steps": 1360, "total_steps": 2322, "loss": 0.31598858833312987, "lr": 3.6765244304380323e-06, "epoch": 1.1714039621016366, "percentage": 58.57, "elapsed_time": "5:12:28", "remaining_time": "3:41:01", "throughput": 7414.05, "total_tokens": 139004040}
+{"current_steps": 1365, "total_steps": 2322, "loss": 0.31764779090881345, "lr": 3.643937188797826e-06, "epoch": 1.1757105943152455, "percentage": 58.79, "elapsed_time": "5:13:33", "remaining_time": "3:39:49", "throughput": 7416.53, "total_tokens": 139527672}
+{"current_steps": 1370, "total_steps": 2322, "loss": 0.30363523960113525, "lr": 3.611412004542887e-06, "epoch": 1.1800172265288544, "percentage": 59.0, "elapsed_time": "5:14:42", "remaining_time": "3:38:40", "throughput": 7417.18, "total_tokens": 140051536}
+{"current_steps": 1375, "total_steps": 2322, "loss": 0.3027554750442505, "lr": 3.578950366120414e-06, "epoch": 1.1843238587424634, "percentage": 59.22, "elapsed_time": "5:15:39", "remaining_time": "3:37:24", "throughput": 7419.65, "total_tokens": 140523744}
+{"current_steps": 1380, "total_steps": 2322, "loss": 0.27700581550598147, "lr": 3.54655375906956e-06, "epoch": 1.1886304909560723, "percentage": 59.43, "elapsed_time": "5:16:55", "remaining_time": "3:36:20", "throughput": 7414.56, "total_tokens": 140991200}
+{"current_steps": 1385, "total_steps": 2322, "loss": 0.3062134742736816, "lr": 3.514223665953455e-06, "epoch": 1.1929371231696813, "percentage": 59.65, "elapsed_time": "5:18:05", "remaining_time": "3:35:11", "throughput": 7415.21, "total_tokens": 141521152}
+{"current_steps": 1390, "total_steps": 2322, "loss": 0.2952254772186279, "lr": 3.481961566291358e-06, "epoch": 1.1972437553832902, "percentage": 59.86, "elapsed_time": "5:19:13", "remaining_time": "3:34:02", "throughput": 7416.49, "total_tokens": 142052768}
+{"current_steps": 1395, "total_steps": 2322, "loss": 0.30109152793884275, "lr": 3.4497689364909483e-06, "epoch": 1.2015503875968991, "percentage": 60.08, "elapsed_time": "5:20:33", "remaining_time": "3:33:00", "throughput": 7413.07, "total_tokens": 142577656}
+{"current_steps": 1400, "total_steps": 2322, "loss": 0.30313336849212646, "lr": 3.417647249780761e-06, "epoch": 1.2058570198105083, "percentage": 60.29, "elapsed_time": "5:21:29", "remaining_time": "3:31:43", "throughput": 7418.22, "total_tokens": 143091504}
+{"current_steps": 1405, "total_steps": 2322, "loss": 0.30561926364898684, "lr": 3.3855979761427705e-06, "epoch": 1.210163652024117, "percentage": 60.51, "elapsed_time": "5:22:32", "remaining_time": "3:30:30", "throughput": 7419.19, "total_tokens": 143579880}
+{"current_steps": 1410, "total_steps": 2322, "loss": 0.3229517459869385, "lr": 3.3536225822451164e-06, "epoch": 1.2144702842377262, "percentage": 60.72, "elapsed_time": "5:23:39", "remaining_time": "3:29:20", "throughput": 7421.33, "total_tokens": 144116304}
+{"current_steps": 1415, "total_steps": 2322, "loss": 0.3148001432418823, "lr": 3.321722531374988e-06, "epoch": 1.2187769164513351, "percentage": 60.94, "elapsed_time": "5:24:57", "remaining_time": "3:28:17", "throughput": 7416.78, "total_tokens": 144609456}
+{"current_steps": 1420, "total_steps": 2322, "loss": 0.3218817710876465, "lr": 3.289899283371657e-06, "epoch": 1.223083548664944, "percentage": 61.15, "elapsed_time": "5:26:02", "remaining_time": "3:27:06", "throughput": 7416.58, "total_tokens": 145088128}
+{"current_steps": 1425, "total_steps": 2322, "loss": 0.3102770566940308, "lr": 3.258154294559671e-06, "epoch": 1.227390180878553, "percentage": 61.37, "elapsed_time": "5:27:13", "remaining_time": "3:25:59", "throughput": 7415.18, "total_tokens": 145589104}
+{"current_steps": 1430, "total_steps": 2322, "loss": 0.28916454315185547, "lr": 3.2264890176822137e-06, "epoch": 1.231696813092162, "percentage": 61.58, "elapsed_time": "5:28:24", "remaining_time": "3:24:51", "throughput": 7413.81, "total_tokens": 146084280}
+{"current_steps": 1435, "total_steps": 2322, "loss": 0.2934903860092163, "lr": 3.194904901834613e-06, "epoch": 1.2360034453057709, "percentage": 61.8, "elapsed_time": "5:29:26", "remaining_time": "3:23:37", "throughput": 7415.75, "total_tokens": 146581160}
+{"current_steps": 1440, "total_steps": 2322, "loss": 0.2953122854232788, "lr": 3.1634033923980336e-06, "epoch": 1.2403100775193798, "percentage": 62.02, "elapsed_time": "5:30:29", "remaining_time": "3:22:25", "throughput": 7416.97, "total_tokens": 147076232}
+{"current_steps": 1445, "total_steps": 2322, "loss": 0.3041204929351807, "lr": 3.131985930973329e-06, "epoch": 1.2446167097329888, "percentage": 62.23, "elapsed_time": "5:31:34", "remaining_time": "3:21:14", "throughput": 7418.47, "total_tokens": 147584128}
+{"current_steps": 1450, "total_steps": 2322, "loss": 0.3094919204711914, "lr": 3.1006539553150727e-06, "epoch": 1.2489233419465977, "percentage": 62.45, "elapsed_time": "5:32:51", "remaining_time": "3:20:10", "throughput": 7417.08, "total_tokens": 148127248}
+{"current_steps": 1455, "total_steps": 2322, "loss": 0.31122531890869143, "lr": 3.0694088992657617e-06, "epoch": 1.2532299741602067, "percentage": 62.66, "elapsed_time": "5:33:57", "remaining_time": "3:19:00", "throughput": 7417.22, "total_tokens": 148624776}
+{"current_steps": 1460, "total_steps": 2322, "loss": 0.30587000846862794, "lr": 3.0382521926901946e-06, "epoch": 1.2575366063738156, "percentage": 62.88, "elapsed_time": "5:35:01", "remaining_time": "3:17:48", "throughput": 7416.55, "total_tokens": 149084784}
+{"current_steps": 1465, "total_steps": 2322, "loss": 0.3310825347900391, "lr": 3.0071852614100427e-06, "epoch": 1.2618432385874248, "percentage": 63.09, "elapsed_time": "5:36:15", "remaining_time": "3:16:42", "throughput": 7415.52, "total_tokens": 149611248}
+{"current_steps": 1470, "total_steps": 2322, "loss": 0.31157741546630857, "lr": 2.9762095271385995e-06, "epoch": 1.2661498708010335, "percentage": 63.31, "elapsed_time": "5:37:25", "remaining_time": "3:15:34", "throughput": 7414.52, "total_tokens": 150113280}
+{"current_steps": 1475, "total_steps": 2322, "loss": 0.29090151786804197, "lr": 2.9453264074157134e-06, "epoch": 1.2704565030146426, "percentage": 63.52, "elapsed_time": "5:38:37", "remaining_time": "3:14:26", "throughput": 7411.45, "total_tokens": 150580240}
+{"current_steps": 1480, "total_steps": 2322, "loss": 0.3190771102905273, "lr": 2.9145373155429263e-06, "epoch": 1.2747631352282516, "percentage": 63.74, "elapsed_time": "5:39:40", "remaining_time": "3:13:15", "throughput": 7413.66, "total_tokens": 151097168}
+{"current_steps": 1485, "total_steps": 2322, "loss": 0.31575181484222414, "lr": 2.883843660518787e-06, "epoch": 1.2790697674418605, "percentage": 63.95, "elapsed_time": "5:40:45", "remaining_time": "3:12:03", "throughput": 7415.83, "total_tokens": 151620032}
+{"current_steps": 1490, "total_steps": 2322, "loss": 0.31672372817993166, "lr": 2.853246846974382e-06, "epoch": 1.2833763996554695, "percentage": 64.17, "elapsed_time": "5:42:02", "remaining_time": "3:10:59", "throughput": 7414.47, "total_tokens": 152160432}
+{"current_steps": 1495, "total_steps": 2322, "loss": 0.3057991027832031, "lr": 2.8227482751090445e-06, "epoch": 1.2876830318690784, "percentage": 64.38, "elapsed_time": "5:42:56", "remaining_time": "3:09:42", "throughput": 7417.8, "total_tokens": 152629192}
+{"current_steps": 1500, "total_steps": 2322, "loss": 0.3172436237335205, "lr": 2.7923493406262836e-06, "epoch": 1.2919896640826873, "percentage": 64.6, "elapsed_time": "5:44:05", "remaining_time": "3:08:33", "throughput": 7417.69, "total_tokens": 153140224}
+{"current_steps": 1505, "total_steps": 2322, "loss": 0.2983067512512207, "lr": 2.7620514346699103e-06, "epoch": 1.2962962962962963, "percentage": 64.81, "elapsed_time": "5:45:19", "remaining_time": "3:07:27", "throughput": 7416.09, "total_tokens": 153656144}
+{"current_steps": 1510, "total_steps": 2322, "loss": 0.31225831508636476, "lr": 2.73185594376038e-06, "epoch": 1.3006029285099052, "percentage": 65.03, "elapsed_time": "5:46:26", "remaining_time": "3:06:17", "throughput": 7416.19, "total_tokens": 154153816}
+{"current_steps": 1515, "total_steps": 2322, "loss": 0.302036190032959, "lr": 2.7017642497313324e-06, "epoch": 1.3049095607235142, "percentage": 65.25, "elapsed_time": "5:47:34", "remaining_time": "3:05:08", "throughput": 7416.69, "total_tokens": 154670920}
+{"current_steps": 1520, "total_steps": 2322, "loss": 0.3167644262313843, "lr": 2.671777729666363e-06, "epoch": 1.309216192937123, "percentage": 65.46, "elapsed_time": "5:48:50", "remaining_time": "3:04:03", "throughput": 7414.6, "total_tokens": 155192016}
+{"current_steps": 1525, "total_steps": 2322, "loss": 0.30224175453186036, "lr": 2.641897755835997e-06, "epoch": 1.313522825150732, "percentage": 65.68, "elapsed_time": "5:50:04", "remaining_time": "3:02:57", "throughput": 7412.97, "total_tokens": 155707640}
+{"current_steps": 1530, "total_steps": 2322, "loss": 0.31644363403320314, "lr": 2.6121256956348966e-06, "epoch": 1.3178294573643412, "percentage": 65.89, "elapsed_time": "5:51:19", "remaining_time": "3:01:51", "throughput": 7412.6, "total_tokens": 156250728}
+{"current_steps": 1535, "total_steps": 2322, "loss": 0.3076641082763672, "lr": 2.58246291151928e-06, "epoch": 1.32213608957795, "percentage": 66.11, "elapsed_time": "5:52:31", "remaining_time": "3:00:44", "throughput": 7410.67, "total_tokens": 156749112}
+{"current_steps": 1540, "total_steps": 2322, "loss": 0.3091968774795532, "lr": 2.5529107609445737e-06, "epoch": 1.326442721791559, "percentage": 66.32, "elapsed_time": "5:53:43", "remaining_time": "2:59:36", "throughput": 7409.8, "total_tokens": 157258624}
+{"current_steps": 1545, "total_steps": 2322, "loss": 0.31270341873168944, "lr": 2.5234705963032917e-06, "epoch": 1.330749354005168, "percentage": 66.54, "elapsed_time": "5:54:53", "remaining_time": "2:58:28", "throughput": 7407.8, "total_tokens": 157737392}
+{"current_steps": 1550, "total_steps": 2322, "loss": 0.29509544372558594, "lr": 2.494143764863142e-06, "epoch": 1.335055986218777, "percentage": 66.75, "elapsed_time": "5:55:57", "remaining_time": "2:57:17", "throughput": 7408.18, "total_tokens": 158223688}
+{"current_steps": 1555, "total_steps": 2322, "loss": 0.28979308605194093, "lr": 2.464931608705384e-06, "epoch": 1.339362618432386, "percentage": 66.97, "elapsed_time": "5:56:55", "remaining_time": "2:56:03", "throughput": 7409.95, "total_tokens": 158688096}
+{"current_steps": 1560, "total_steps": 2322, "loss": 0.3088102102279663, "lr": 2.4358354646633915e-06, "epoch": 1.3436692506459949, "percentage": 67.18, "elapsed_time": "5:58:03", "remaining_time": "2:54:53", "throughput": 7409.83, "total_tokens": 159187424}
+{"current_steps": 1565, "total_steps": 2322, "loss": 0.29939565658569334, "lr": 2.4068566642614923e-06, "epoch": 1.3479758828596038, "percentage": 67.4, "elapsed_time": "5:59:10", "remaining_time": "2:53:43", "throughput": 7409.22, "total_tokens": 159669688}
+{"current_steps": 1570, "total_steps": 2322, "loss": 0.3042983770370483, "lr": 2.3779965336540237e-06, "epoch": 1.3522825150732127, "percentage": 67.61, "elapsed_time": "6:00:11", "remaining_time": "2:52:31", "throughput": 7411.39, "total_tokens": 160172312}
+{"current_steps": 1575, "total_steps": 2322, "loss": 0.2935898542404175, "lr": 2.3492563935646493e-06, "epoch": 1.3565891472868217, "percentage": 67.83, "elapsed_time": "6:01:28", "remaining_time": "2:51:26", "throughput": 7409.97, "total_tokens": 160708952}
+{"current_steps": 1580, "total_steps": 2322, "loss": 0.30011003017425536, "lr": 2.3206375592259205e-06, "epoch": 1.3608957795004306, "percentage": 68.04, "elapsed_time": "6:02:41", "remaining_time": "2:50:19", "throughput": 7408.5, "total_tokens": 161217216}
+{"current_steps": 1585, "total_steps": 2322, "loss": 0.3024036645889282, "lr": 2.2921413403190774e-06, "epoch": 1.3652024117140396, "percentage": 68.26, "elapsed_time": "6:03:56", "remaining_time": "2:49:13", "throughput": 7405.34, "total_tokens": 161705312}
+{"current_steps": 1590, "total_steps": 2322, "loss": 0.30027027130126954, "lr": 2.263769040914125e-06, "epoch": 1.3695090439276485, "percentage": 68.48, "elapsed_time": "6:04:56", "remaining_time": "2:48:00", "throughput": 7405.86, "total_tokens": 162163072}
+{"current_steps": 1595, "total_steps": 2322, "loss": 0.3255805492401123, "lr": 2.2355219594101483e-06, "epoch": 1.3738156761412577, "percentage": 68.69, "elapsed_time": "6:06:06", "remaining_time": "2:46:52", "throughput": 7405.74, "total_tokens": 162678312}
+{"current_steps": 1600, "total_steps": 2322, "loss": 0.31427993774414065, "lr": 2.2074013884758993e-06, "epoch": 1.3781223083548664, "percentage": 68.91, "elapsed_time": "6:07:12", "remaining_time": "2:45:42", "throughput": 7408.19, "total_tokens": 163224632}
+{"current_steps": 1605, "total_steps": 2322, "loss": 0.3022661924362183, "lr": 2.179408614990635e-06, "epoch": 1.3824289405684755, "percentage": 69.12, "elapsed_time": "6:08:19", "remaining_time": "2:44:32", "throughput": 7407.23, "total_tokens": 163699352}
+{"current_steps": 1610, "total_steps": 2322, "loss": 0.31579549312591554, "lr": 2.1515449199852307e-06, "epoch": 1.3867355727820845, "percentage": 69.34, "elapsed_time": "6:09:30", "remaining_time": "2:43:24", "throughput": 7407.96, "total_tokens": 164236200}
+{"current_steps": 1615, "total_steps": 2322, "loss": 0.31688437461853025, "lr": 2.1238115785835512e-06, "epoch": 1.3910422049956934, "percentage": 69.55, "elapsed_time": "6:10:41", "remaining_time": "2:42:16", "throughput": 7406.65, "total_tokens": 164733432}
+{"current_steps": 1620, "total_steps": 2322, "loss": 0.3063546657562256, "lr": 2.0962098599441065e-06, "epoch": 1.3953488372093024, "percentage": 69.77, "elapsed_time": "6:11:44", "remaining_time": "2:41:05", "throughput": 7408.18, "total_tokens": 165233040}
+{"current_steps": 1625, "total_steps": 2322, "loss": 0.28777213096618653, "lr": 2.068741027201961e-06, "epoch": 1.3996554694229113, "percentage": 69.98, "elapsed_time": "6:12:51", "remaining_time": "2:39:55", "throughput": 7406.9, "total_tokens": 165704392}
+{"current_steps": 1630, "total_steps": 2322, "loss": 0.3152973413467407, "lr": 2.0414063374109326e-06, "epoch": 1.4039621016365202, "percentage": 70.2, "elapsed_time": "6:13:59", "remaining_time": "2:38:46", "throughput": 7407.33, "total_tokens": 166213024}
+{"current_steps": 1635, "total_steps": 2322, "loss": 0.29006712436676024, "lr": 2.0142070414860704e-06, "epoch": 1.4082687338501292, "percentage": 70.41, "elapsed_time": "6:14:58", "remaining_time": "2:37:33", "throughput": 7408.83, "total_tokens": 166683888}
+{"current_steps": 1640, "total_steps": 2322, "loss": 0.3009150505065918, "lr": 1.9871443841464004e-06, "epoch": 1.4125753660637381, "percentage": 70.63, "elapsed_time": "6:16:10", "remaining_time": "2:36:25", "throughput": 7408.33, "total_tokens": 167208664}
+{"current_steps": 1645, "total_steps": 2322, "loss": 0.308090877532959, "lr": 1.9602196038579774e-06, "epoch": 1.416881998277347, "percentage": 70.84, "elapsed_time": "6:17:16", "remaining_time": "2:35:16", "throughput": 7408.92, "total_tokens": 167711896}
+{"current_steps": 1650, "total_steps": 2322, "loss": 0.31527588367462156, "lr": 1.933433932777193e-06, "epoch": 1.421188630490956, "percentage": 71.06, "elapsed_time": "6:18:30", "remaining_time": "2:34:09", "throughput": 7408.78, "total_tokens": 168253824}
+{"current_steps": 1655, "total_steps": 2322, "loss": 0.2953861474990845, "lr": 1.9067885966943983e-06, "epoch": 1.425495262704565, "percentage": 71.27, "elapsed_time": "6:19:38", "remaining_time": "2:33:00", "throughput": 7408.51, "total_tokens": 168753216}
+{"current_steps": 1660, "total_steps": 2322, "loss": 0.3124719142913818, "lr": 1.880284814977807e-06, "epoch": 1.429801894918174, "percentage": 71.49, "elapsed_time": "6:20:44", "remaining_time": "2:31:50", "throughput": 7408.31, "total_tokens": 169237536}
+{"current_steps": 1665, "total_steps": 2322, "loss": 0.28069252967834474, "lr": 1.8539238005176912e-06, "epoch": 1.4341085271317828, "percentage": 71.71, "elapsed_time": "6:21:57", "remaining_time": "2:30:43", "throughput": 7406.55, "total_tokens": 169742192}
+{"current_steps": 1670, "total_steps": 2322, "loss": 0.30810205936431884, "lr": 1.8277067596708793e-06, "epoch": 1.438415159345392, "percentage": 71.92, "elapsed_time": "6:23:07", "remaining_time": "2:29:34", "throughput": 7404.99, "total_tokens": 170221760}
+{"current_steps": 1675, "total_steps": 2322, "loss": 0.3010256767272949, "lr": 1.8016348922055448e-06, "epoch": 1.442721791559001, "percentage": 72.14, "elapsed_time": "6:24:16", "remaining_time": "2:28:26", "throughput": 7404.33, "total_tokens": 170720016}
+{"current_steps": 1680, "total_steps": 2322, "loss": 0.30887162685394287, "lr": 1.7757093912463058e-06, "epoch": 1.4470284237726099, "percentage": 72.35, "elapsed_time": "6:25:23", "remaining_time": "2:27:16", "throughput": 7405.14, "total_tokens": 171232760}
+{"current_steps": 1685, "total_steps": 2322, "loss": 0.3016911506652832, "lr": 1.7499314432196257e-06, "epoch": 1.4513350559862188, "percentage": 72.57, "elapsed_time": "6:26:36", "remaining_time": "2:26:09", "throughput": 7403.48, "total_tokens": 171731496}
+{"current_steps": 1690, "total_steps": 2322, "loss": 0.3081092357635498, "lr": 1.7243022277995109e-06, "epoch": 1.4556416881998278, "percentage": 72.78, "elapsed_time": "6:27:50", "remaining_time": "2:25:02", "throughput": 7402.04, "total_tokens": 172245912}
+{"current_steps": 1695, "total_steps": 2322, "loss": 0.3349776268005371, "lr": 1.698822917853532e-06, "epoch": 1.4599483204134367, "percentage": 73.0, "elapsed_time": "6:28:58", "remaining_time": "2:23:53", "throughput": 7403.26, "total_tokens": 172780776}
+{"current_steps": 1700, "total_steps": 2322, "loss": 0.2948931694030762, "lr": 1.6734946793891498e-06, "epoch": 1.4642549526270456, "percentage": 73.21, "elapsed_time": "6:29:53", "remaining_time": "2:22:39", "throughput": 7406.28, "total_tokens": 173260968}
+{"current_steps": 1705, "total_steps": 2322, "loss": 0.3255448818206787, "lr": 1.6483186715003523e-06, "epoch": 1.4685615848406546, "percentage": 73.43, "elapsed_time": "6:30:53", "remaining_time": "2:21:27", "throughput": 7409.23, "total_tokens": 173774184}
+{"current_steps": 1710, "total_steps": 2322, "loss": 0.3152692556381226, "lr": 1.623296046314612e-06, "epoch": 1.4728682170542635, "percentage": 73.64, "elapsed_time": "6:32:05", "remaining_time": "2:20:19", "throughput": 7408.94, "total_tokens": 174299296}
+{"current_steps": 1715, "total_steps": 2322, "loss": 0.3075227737426758, "lr": 1.5984279489401655e-06, "epoch": 1.4771748492678725, "percentage": 73.86, "elapsed_time": "6:33:19", "remaining_time": "2:19:12", "throughput": 7406.87, "total_tokens": 174801056}
+{"current_steps": 1720, "total_steps": 2322, "loss": 0.2994909524917603, "lr": 1.5737155174136042e-06, "epoch": 1.4814814814814814, "percentage": 74.07, "elapsed_time": "6:34:29", "remaining_time": "2:18:04", "throughput": 7406.76, "total_tokens": 175312064}
+{"current_steps": 1725, "total_steps": 2322, "loss": 0.2942391872406006, "lr": 1.5491598826477967e-06, "epoch": 1.4857881136950906, "percentage": 74.29, "elapsed_time": "6:35:35", "remaining_time": "2:16:54", "throughput": 7407.05, "total_tokens": 175813120}
+{"current_steps": 1730, "total_steps": 2322, "loss": 0.30359663963317873, "lr": 1.5247621683801344e-06, "epoch": 1.4900947459086993, "percentage": 74.5, "elapsed_time": "6:36:53", "remaining_time": "2:15:48", "throughput": 7403.96, "total_tokens": 176311208}
+{"current_steps": 1735, "total_steps": 2322, "loss": 0.30945577621459963, "lr": 1.500523491121108e-06, "epoch": 1.4944013781223084, "percentage": 74.72, "elapsed_time": "6:37:55", "remaining_time": "2:14:37", "throughput": 7406.53, "total_tokens": 176837256}
+{"current_steps": 1740, "total_steps": 2322, "loss": 0.3017910957336426, "lr": 1.476444960103211e-06, "epoch": 1.4987080103359174, "percentage": 74.94, "elapsed_time": "6:39:10", "remaining_time": "2:13:30", "throughput": 7404.63, "total_tokens": 177342432}
+{"current_steps": 1745, "total_steps": 2322, "loss": 0.2982076644897461, "lr": 1.4525276772301761e-06, "epoch": 1.5030146425495263, "percentage": 75.15, "elapsed_time": "6:40:22", "remaining_time": "2:12:23", "throughput": 7403.01, "total_tokens": 177837496}
+{"current_steps": 1750, "total_steps": 2322, "loss": 0.305989408493042, "lr": 1.4287727370265558e-06, "epoch": 1.5073212747631353, "percentage": 75.37, "elapsed_time": "6:41:38", "remaining_time": "2:11:16", "throughput": 7401.75, "total_tokens": 178367400}
+{"current_steps": 1755, "total_steps": 2322, "loss": 0.31029796600341797, "lr": 1.4051812265876257e-06, "epoch": 1.5116279069767442, "percentage": 75.58, "elapsed_time": "6:42:48", "remaining_time": "2:10:08", "throughput": 7401.4, "total_tokens": 178879808}
+{"current_steps": 1760, "total_steps": 2322, "loss": 0.29440352916717527, "lr": 1.381754225529639e-06, "epoch": 1.5159345391903531, "percentage": 75.8, "elapsed_time": "6:43:55", "remaining_time": "2:08:58", "throughput": 7402.3, "total_tokens": 179399168}
+{"current_steps": 1765, "total_steps": 2322, "loss": 0.3081679344177246, "lr": 1.3584928059404207e-06, "epoch": 1.520241171403962, "percentage": 76.01, "elapsed_time": "6:45:01", "remaining_time": "2:07:49", "throughput": 7403.29, "total_tokens": 179913984}
+{"current_steps": 1770, "total_steps": 2322, "loss": 0.28179948329925536, "lr": 1.3353980323303067e-06, "epoch": 1.524547803617571, "percentage": 76.23, "elapsed_time": "6:46:04", "remaining_time": "2:06:38", "throughput": 7403.52, "total_tokens": 180380992}
+{"current_steps": 1775, "total_steps": 2322, "loss": 0.3199465274810791, "lr": 1.3124709615834263e-06, "epoch": 1.52885443583118, "percentage": 76.44, "elapsed_time": "6:47:08", "remaining_time": "2:05:28", "throughput": 7405.19, "total_tokens": 180895408}
+{"current_steps": 1780, "total_steps": 2322, "loss": 0.31137895584106445, "lr": 1.2897126429093354e-06, "epoch": 1.5331610680447891, "percentage": 76.66, "elapsed_time": "6:48:21", "remaining_time": "2:04:20", "throughput": 7403.88, "total_tokens": 181405216}
+{"current_steps": 1785, "total_steps": 2322, "loss": 0.2975735902786255, "lr": 1.2671241177950078e-06, "epoch": 1.5374677002583979, "percentage": 76.87, "elapsed_time": "6:49:34", "remaining_time": "2:03:12", "throughput": 7403.26, "total_tokens": 181929944}
+{"current_steps": 1790, "total_steps": 2322, "loss": 0.30163776874542236, "lr": 1.2447064199571657e-06, "epoch": 1.541774332472007, "percentage": 77.09, "elapsed_time": "6:50:47", "remaining_time": "2:02:05", "throughput": 7402.96, "total_tokens": 182464632}
+{"current_steps": 1795, "total_steps": 2322, "loss": 0.2984232664108276, "lr": 1.2224605752949786e-06, "epoch": 1.5460809646856157, "percentage": 77.3, "elapsed_time": "6:51:48", "remaining_time": "2:00:54", "throughput": 7404.83, "total_tokens": 182958712}
+{"current_steps": 1800, "total_steps": 2322, "loss": 0.29271659851074217, "lr": 1.2003876018431133e-06, "epoch": 1.550387596899225, "percentage": 77.52, "elapsed_time": "6:52:40", "remaining_time": "1:59:40", "throughput": 7408.47, "total_tokens": 183435376}
+{"current_steps": 1805, "total_steps": 2322, "loss": 0.2989810466766357, "lr": 1.1784885097251474e-06, "epoch": 1.5546942291128336, "percentage": 77.73, "elapsed_time": "6:53:36", "remaining_time": "1:58:28", "throughput": 7412.03, "total_tokens": 183939080}
+{"current_steps": 1810, "total_steps": 2322, "loss": 0.31319284439086914, "lr": 1.1567643011073393e-06, "epoch": 1.5590008613264428, "percentage": 77.95, "elapsed_time": "6:54:47", "remaining_time": "1:57:19", "throughput": 7413.2, "total_tokens": 184492224}
+{"current_steps": 1815, "total_steps": 2322, "loss": 0.3045823574066162, "lr": 1.1352159701527743e-06, "epoch": 1.5633074935400517, "percentage": 78.17, "elapsed_time": "6:56:05", "remaining_time": "1:56:13", "throughput": 7410.83, "total_tokens": 185017248}
+{"current_steps": 1820, "total_steps": 2322, "loss": 0.29555253982543944, "lr": 1.1138445029758594e-06, "epoch": 1.5676141257536607, "percentage": 78.38, "elapsed_time": "6:57:19", "remaining_time": "1:55:06", "throughput": 7409.45, "total_tokens": 185525392}
+{"current_steps": 1825, "total_steps": 2322, "loss": 0.29001967906951903, "lr": 1.0926508775971995e-06, "epoch": 1.5719207579672696, "percentage": 78.6, "elapsed_time": "6:58:28", "remaining_time": "1:53:57", "throughput": 7407.95, "total_tokens": 186004920}
+{"current_steps": 1830, "total_steps": 2322, "loss": 0.2898190259933472, "lr": 1.0716360638988416e-06, "epoch": 1.5762273901808785, "percentage": 78.81, "elapsed_time": "6:59:33", "remaining_time": "1:52:47", "throughput": 7409.74, "total_tokens": 186528776}
+{"current_steps": 1835, "total_steps": 2322, "loss": 0.3022593975067139, "lr": 1.0508010235798904e-06, "epoch": 1.5805340223944875, "percentage": 79.03, "elapsed_time": "7:00:37", "remaining_time": "1:51:37", "throughput": 7410.8, "total_tokens": 187030096}
+{"current_steps": 1840, "total_steps": 2322, "loss": 0.29773921966552735, "lr": 1.0301467101124956e-06, "epoch": 1.5848406546080964, "percentage": 79.24, "elapsed_time": "7:01:40", "remaining_time": "1:50:27", "throughput": 7412.24, "total_tokens": 187530704}
+{"current_steps": 1845, "total_steps": 2322, "loss": 0.2982914447784424, "lr": 1.0096740686982192e-06, "epoch": 1.5891472868217056, "percentage": 79.46, "elapsed_time": "7:02:54", "remaining_time": "1:49:20", "throughput": 7408.83, "total_tokens": 187997312}
+{"current_steps": 1850, "total_steps": 2322, "loss": 0.292753005027771, "lr": 9.893840362247809e-07, "epoch": 1.5934539190353143, "percentage": 79.67, "elapsed_time": "7:04:03", "remaining_time": "1:48:11", "throughput": 7408.1, "total_tokens": 188486784}
+{"current_steps": 1855, "total_steps": 2322, "loss": 0.3059796571731567, "lr": 9.692775412231863e-07, "epoch": 1.5977605512489235, "percentage": 79.89, "elapsed_time": "7:05:18", "remaining_time": "1:47:04", "throughput": 7406.48, "total_tokens": 189005680}
+{"current_steps": 1860, "total_steps": 2322, "loss": 0.29838857650756834, "lr": 9.493555038252273e-07, "epoch": 1.6020671834625322, "percentage": 80.1, "elapsed_time": "7:06:23", "remaining_time": "1:45:54", "throughput": 7407.58, "total_tokens": 189510720}
+{"current_steps": 1865, "total_steps": 2322, "loss": 0.3135526180267334, "lr": 9.296188357213804e-07, "epoch": 1.6063738156761413, "percentage": 80.32, "elapsed_time": "7:07:35", "remaining_time": "1:44:46", "throughput": 7406.1, "total_tokens": 190005232}
+{"current_steps": 1870, "total_steps": 2322, "loss": 0.2981297492980957, "lr": 9.100684401190829e-07, "epoch": 1.61068044788975, "percentage": 80.53, "elapsed_time": "7:08:49", "remaining_time": "1:43:39", "throughput": 7404.6, "total_tokens": 190519568}
+{"current_steps": 1875, "total_steps": 2322, "loss": 0.3049411058425903, "lr": 8.907052117013981e-07, "epoch": 1.6149870801033592, "percentage": 80.75, "elapsed_time": "7:10:04", "remaining_time": "1:42:31", "throughput": 7403.67, "total_tokens": 191051376}
+{"current_steps": 1880, "total_steps": 2322, "loss": 0.2883660316467285, "lr": 8.715300365860784e-07, "epoch": 1.6192937123169682, "percentage": 80.96, "elapsed_time": "7:11:12", "remaining_time": "1:41:22", "throughput": 7403.7, "total_tokens": 191555640}
+{"current_steps": 1885, "total_steps": 2322, "loss": 0.2918602228164673, "lr": 8.525437922850033e-07, "epoch": 1.623600344530577, "percentage": 81.18, "elapsed_time": "7:12:20", "remaining_time": "1:40:13", "throughput": 7402.97, "total_tokens": 192040232}
+{"current_steps": 1890, "total_steps": 2322, "loss": 0.296920108795166, "lr": 8.337473476640318e-07, "epoch": 1.627906976744186, "percentage": 81.4, "elapsed_time": "7:13:19", "remaining_time": "1:39:02", "throughput": 7406.14, "total_tokens": 192555328}
+{"current_steps": 1895, "total_steps": 2322, "loss": 0.3084681749343872, "lr": 8.151415629032338e-07, "epoch": 1.632213608957795, "percentage": 81.61, "elapsed_time": "7:14:31", "remaining_time": "1:37:54", "throughput": 7404.49, "total_tokens": 193047904}
+{"current_steps": 1900, "total_steps": 2322, "loss": 0.3049050807952881, "lr": 7.967272894575312e-07, "epoch": 1.636520241171404, "percentage": 81.83, "elapsed_time": "7:15:42", "remaining_time": "1:36:46", "throughput": 7404.83, "total_tokens": 193578488}
+{"current_steps": 1905, "total_steps": 2322, "loss": 0.31264266967773435, "lr": 7.785053700177275e-07, "epoch": 1.6408268733850129, "percentage": 82.04, "elapsed_time": "7:16:47", "remaining_time": "1:35:36", "throughput": 7406.63, "total_tokens": 194106056}
+{"current_steps": 1910, "total_steps": 2322, "loss": 0.32028186321258545, "lr": 7.604766384719487e-07, "epoch": 1.645133505598622, "percentage": 82.26, "elapsed_time": "7:17:51", "remaining_time": "1:34:26", "throughput": 7408.77, "total_tokens": 194637600}
+{"current_steps": 1915, "total_steps": 2322, "loss": 0.29713547229766846, "lr": 7.426419198674773e-07, "epoch": 1.6494401378122308, "percentage": 82.47, "elapsed_time": "7:19:04", "remaining_time": "1:33:19", "throughput": 7407.14, "total_tokens": 195135888}
+{"current_steps": 1920, "total_steps": 2322, "loss": 0.318635368347168, "lr": 7.250020303730015e-07, "epoch": 1.65374677002584, "percentage": 82.69, "elapsed_time": "7:20:10", "remaining_time": "1:32:09", "throughput": 7408.23, "total_tokens": 195658344}
+{"current_steps": 1925, "total_steps": 2322, "loss": 0.29038522243499754, "lr": 7.075577772412607e-07, "epoch": 1.6580534022394486, "percentage": 82.9, "elapsed_time": "7:21:15", "remaining_time": "1:31:00", "throughput": 7408.98, "total_tokens": 196159872}
+{"current_steps": 1930, "total_steps": 2322, "loss": 0.3107494831085205, "lr": 6.903099587721024e-07, "epoch": 1.6623600344530578, "percentage": 83.12, "elapsed_time": "7:22:23", "remaining_time": "1:29:51", "throughput": 7407.34, "total_tokens": 196617976}
+{"current_steps": 1935, "total_steps": 2322, "loss": 0.29712848663330077, "lr": 6.732593642759533e-07, "epoch": 1.6666666666666665, "percentage": 83.33, "elapsed_time": "7:23:33", "remaining_time": "1:28:42", "throughput": 7406.13, "total_tokens": 197100472}
+{"current_steps": 1940, "total_steps": 2322, "loss": 0.29893629550933837, "lr": 6.564067740376951e-07, "epoch": 1.6709732988802757, "percentage": 83.55, "elapsed_time": "7:24:39", "remaining_time": "1:27:33", "throughput": 7407.34, "total_tokens": 197626872}
+{"current_steps": 1945, "total_steps": 2322, "loss": 0.31168351173400877, "lr": 6.397529592809615e-07, "epoch": 1.6752799310938846, "percentage": 83.76, "elapsed_time": "7:25:46", "remaining_time": "1:26:24", "throughput": 7407.4, "total_tokens": 198123984}
+{"current_steps": 1950, "total_steps": 2322, "loss": 0.29562864303588865, "lr": 6.232986821328374e-07, "epoch": 1.6795865633074936, "percentage": 83.98, "elapsed_time": "7:27:03", "remaining_time": "1:25:17", "throughput": 7406.0, "total_tokens": 198655680}
+{"current_steps": 1955, "total_steps": 2322, "loss": 0.2916395664215088, "lr": 6.070446955889853e-07, "epoch": 1.6838931955211025, "percentage": 84.19, "elapsed_time": "7:28:16", "remaining_time": "1:24:09", "throughput": 7404.94, "total_tokens": 199167544}
+{"current_steps": 1960, "total_steps": 2322, "loss": 0.30132670402526857, "lr": 5.909917434791884e-07, "epoch": 1.6881998277347114, "percentage": 84.41, "elapsed_time": "7:29:23", "remaining_time": "1:22:59", "throughput": 7405.44, "total_tokens": 199673096}
+{"current_steps": 1965, "total_steps": 2322, "loss": 0.3065808296203613, "lr": 5.75140560433306e-07, "epoch": 1.6925064599483204, "percentage": 84.63, "elapsed_time": "7:30:28", "remaining_time": "1:21:50", "throughput": 7406.35, "total_tokens": 200183744}
+{"current_steps": 1970, "total_steps": 2322, "loss": 0.30456061363220216, "lr": 5.594918718476589e-07, "epoch": 1.6968130921619293, "percentage": 84.84, "elapsed_time": "7:31:41", "remaining_time": "1:20:42", "throughput": 7405.38, "total_tokens": 200697696}
+{"current_steps": 1975, "total_steps": 2322, "loss": 0.3014059066772461, "lr": 5.440463938518304e-07, "epoch": 1.7011197243755385, "percentage": 85.06, "elapsed_time": "7:32:52", "remaining_time": "1:19:34", "throughput": 7404.68, "total_tokens": 201206880}
+{"current_steps": 1980, "total_steps": 2322, "loss": 0.2906249284744263, "lr": 5.288048332758961e-07, "epoch": 1.7054263565891472, "percentage": 85.27, "elapsed_time": "7:34:05", "remaining_time": "1:18:25", "throughput": 7402.89, "total_tokens": 201692448}
+{"current_steps": 1985, "total_steps": 2322, "loss": 0.3080711841583252, "lr": 5.137678876180746e-07, "epoch": 1.7097329888027564, "percentage": 85.49, "elapsed_time": "7:35:02", "remaining_time": "1:17:15", "throughput": 7406.22, "total_tokens": 202206200}
+{"current_steps": 1990, "total_steps": 2322, "loss": 0.3220229148864746, "lr": 4.989362450128133e-07, "epoch": 1.714039621016365, "percentage": 85.7, "elapsed_time": "7:36:09", "remaining_time": "1:16:06", "throughput": 7406.61, "total_tokens": 202716392}
+{"current_steps": 1995, "total_steps": 2322, "loss": 0.31665678024291993, "lr": 4.843105841992895e-07, "epoch": 1.7183462532299743, "percentage": 85.92, "elapsed_time": "7:37:08", "remaining_time": "1:14:55", "throughput": 7409.08, "total_tokens": 203222616}
+{"current_steps": 2000, "total_steps": 2322, "loss": 0.2909515380859375, "lr": 4.698915744903559e-07, "epoch": 1.722652885443583, "percentage": 86.13, "elapsed_time": "7:38:23", "remaining_time": "1:13:48", "throughput": 7406.86, "total_tokens": 203712976}
+{"current_steps": 2005, "total_steps": 2322, "loss": 0.299733304977417, "lr": 4.5567987574190677e-07, "epoch": 1.7269595176571921, "percentage": 86.35, "elapsed_time": "7:39:42", "remaining_time": "1:12:40", "throughput": 7403.91, "total_tokens": 204220640}
+{"current_steps": 2010, "total_steps": 2322, "loss": 0.2933858156204224, "lr": 4.416761383226864e-07, "epoch": 1.731266149870801, "percentage": 86.56, "elapsed_time": "7:40:55", "remaining_time": "1:11:32", "throughput": 7403.37, "total_tokens": 204740576}
+{"current_steps": 2015, "total_steps": 2322, "loss": 0.2882672309875488, "lr": 4.278810030845193e-07, "epoch": 1.73557278208441, "percentage": 86.78, "elapsed_time": "7:42:02", "remaining_time": "1:10:23", "throughput": 7404.27, "total_tokens": 205261256}
+{"current_steps": 2020, "total_steps": 2322, "loss": 0.300402307510376, "lr": 4.1429510133298714e-07, "epoch": 1.739879414298019, "percentage": 86.99, "elapsed_time": "7:43:06", "remaining_time": "1:09:14", "throughput": 7405.15, "total_tokens": 205763456}
+{"current_steps": 2025, "total_steps": 2322, "loss": 0.30848591327667235, "lr": 4.0091905479853865e-07, "epoch": 1.744186046511628, "percentage": 87.21, "elapsed_time": "7:44:14", "remaining_time": "1:08:05", "throughput": 7404.6, "total_tokens": 206254632}
+{"current_steps": 2030, "total_steps": 2322, "loss": 0.30381178855895996, "lr": 3.877534756080353e-07, "epoch": 1.7484926787252368, "percentage": 87.42, "elapsed_time": "7:45:25", "remaining_time": "1:06:56", "throughput": 7403.91, "total_tokens": 206761200}
+{"current_steps": 2035, "total_steps": 2322, "loss": 0.30392751693725584, "lr": 3.747989662567403e-07, "epoch": 1.7527993109388458, "percentage": 87.64, "elapsed_time": "7:46:34", "remaining_time": "1:05:48", "throughput": 7403.41, "total_tokens": 207254168}
+{"current_steps": 2040, "total_steps": 2322, "loss": 0.2981263637542725, "lr": 3.620561195807454e-07, "epoch": 1.757105943152455, "percentage": 87.86, "elapsed_time": "7:47:44", "remaining_time": "1:04:39", "throughput": 7401.58, "total_tokens": 207724624}
+{"current_steps": 2045, "total_steps": 2322, "loss": 0.2957895278930664, "lr": 3.4952551872984295e-07, "epoch": 1.7614125753660637, "percentage": 88.07, "elapsed_time": "7:49:00", "remaining_time": "1:03:31", "throughput": 7399.79, "total_tokens": 208230432}
+{"current_steps": 2050, "total_steps": 2322, "loss": 0.31419479846954346, "lr": 3.372077371408361e-07, "epoch": 1.7657192075796728, "percentage": 88.29, "elapsed_time": "7:50:03", "remaining_time": "1:02:22", "throughput": 7401.81, "total_tokens": 208756176}
+{"current_steps": 2055, "total_steps": 2322, "loss": 0.29912557601928713, "lr": 3.2510333851129895e-07, "epoch": 1.7700258397932815, "percentage": 88.5, "elapsed_time": "7:51:16", "remaining_time": "1:01:13", "throughput": 7401.27, "total_tokens": 209283048}
+{"current_steps": 2060, "total_steps": 2322, "loss": 0.2877781391143799, "lr": 3.132128767737824e-07, "epoch": 1.7743324720068907, "percentage": 88.72, "elapsed_time": "7:52:19", "remaining_time": "1:00:04", "throughput": 7402.26, "total_tokens": 209776608}
+{"current_steps": 2065, "total_steps": 2322, "loss": 0.3083071708679199, "lr": 3.015368960704584e-07, "epoch": 1.7786391042204994, "percentage": 88.93, "elapsed_time": "7:53:33", "remaining_time": "0:58:56", "throughput": 7401.15, "total_tokens": 210290720}
+{"current_steps": 2070, "total_steps": 2322, "loss": 0.30655722618103026, "lr": 2.9007593072822246e-07, "epoch": 1.7829457364341086, "percentage": 89.15, "elapsed_time": "7:54:40", "remaining_time": "0:57:47", "throughput": 7401.47, "total_tokens": 210795064}
+{"current_steps": 2075, "total_steps": 2322, "loss": 0.2855481386184692, "lr": 2.7883050523424214e-07, "epoch": 1.7872523686477175, "percentage": 89.36, "elapsed_time": "7:55:50", "remaining_time": "0:56:38", "throughput": 7401.16, "total_tokens": 211307816}
+{"current_steps": 2080, "total_steps": 2322, "loss": 0.3143181800842285, "lr": 2.67801134211953e-07, "epoch": 1.7915590008613265, "percentage": 89.58, "elapsed_time": "7:57:02", "remaining_time": "0:55:30", "throughput": 7400.54, "total_tokens": 211819656}
+{"current_steps": 2085, "total_steps": 2322, "loss": 0.3136578559875488, "lr": 2.569883223975078e-07, "epoch": 1.7958656330749354, "percentage": 89.79, "elapsed_time": "7:58:10", "remaining_time": "0:54:21", "throughput": 7401.22, "total_tokens": 212342848}
+{"current_steps": 2090, "total_steps": 2322, "loss": 0.3073784351348877, "lr": 2.463925646166793e-07, "epoch": 1.8001722652885443, "percentage": 90.01, "elapsed_time": "7:59:24", "remaining_time": "0:53:13", "throughput": 7400.2, "total_tokens": 212865944}
+{"current_steps": 2095, "total_steps": 2322, "loss": 0.2910205364227295, "lr": 2.3601434576221548e-07, "epoch": 1.8044788975021533, "percentage": 90.22, "elapsed_time": "8:00:26", "remaining_time": "0:52:03", "throughput": 7399.8, "total_tokens": 213312424}
+{"current_steps": 2100, "total_steps": 2322, "loss": 0.3124124050140381, "lr": 2.2585414077164914e-07, "epoch": 1.8087855297157622, "percentage": 90.44, "elapsed_time": "8:01:26", "remaining_time": "0:50:53", "throughput": 7403.55, "total_tokens": 213861568}
+{"current_steps": 2105, "total_steps": 2322, "loss": 0.3035388469696045, "lr": 2.1591241460556355e-07, "epoch": 1.8130921619293714, "percentage": 90.65, "elapsed_time": "8:02:42", "remaining_time": "0:49:45", "throughput": 7402.8, "total_tokens": 214400864}
+{"current_steps": 2110, "total_steps": 2322, "loss": 0.30562429428100585, "lr": 2.0618962222631434e-07, "epoch": 1.8173987941429801, "percentage": 90.87, "elapsed_time": "8:03:39", "remaining_time": "0:48:35", "throughput": 7406.48, "total_tokens": 214933176}
+{"current_steps": 2115, "total_steps": 2322, "loss": 0.3033386945724487, "lr": 1.9668620857720865e-07, "epoch": 1.8217054263565893, "percentage": 91.09, "elapsed_time": "8:04:41", "remaining_time": "0:47:26", "throughput": 7406.33, "total_tokens": 215387240}
+{"current_steps": 2120, "total_steps": 2322, "loss": 0.30868115425109866, "lr": 1.8740260856214388e-07, "epoch": 1.826012058570198, "percentage": 91.3, "elapsed_time": "8:05:50", "remaining_time": "0:46:17", "throughput": 7407.7, "total_tokens": 215936824}
+{"current_steps": 2125, "total_steps": 2322, "loss": 0.30875744819641116, "lr": 1.7833924702570725e-07, "epoch": 1.8303186907838072, "percentage": 91.52, "elapsed_time": "8:07:06", "remaining_time": "0:45:09", "throughput": 7407.01, "total_tokens": 216480984}
+{"current_steps": 2130, "total_steps": 2322, "loss": 0.2976231098175049, "lr": 1.6949653873372828e-07, "epoch": 1.8346253229974159, "percentage": 91.73, "elapsed_time": "8:08:23", "remaining_time": "0:44:01", "throughput": 7404.79, "total_tokens": 216985056}
+{"current_steps": 2135, "total_steps": 2322, "loss": 0.3137772560119629, "lr": 1.6087488835430208e-07, "epoch": 1.838931955211025, "percentage": 91.95, "elapsed_time": "8:09:40", "remaining_time": "0:42:53", "throughput": 7405.16, "total_tokens": 217567072}
+{"current_steps": 2140, "total_steps": 2322, "loss": 0.304307746887207, "lr": 1.5247469043927153e-07, "epoch": 1.843238587424634, "percentage": 92.16, "elapsed_time": "8:10:45", "remaining_time": "0:41:44", "throughput": 7405.71, "total_tokens": 218066976}
+{"current_steps": 2145, "total_steps": 2322, "loss": 0.31688547134399414, "lr": 1.4429632940616721e-07, "epoch": 1.847545219638243, "percentage": 92.38, "elapsed_time": "8:11:59", "remaining_time": "0:40:35", "throughput": 7403.62, "total_tokens": 218547776}
+{"current_steps": 2150, "total_steps": 2322, "loss": 0.2871793270111084, "lr": 1.3634017952061872e-07, "epoch": 1.8518518518518519, "percentage": 92.59, "elapsed_time": "8:13:13", "remaining_time": "0:39:27", "throughput": 7403.58, "total_tokens": 219097408}
+{"current_steps": 2155, "total_steps": 2322, "loss": 0.3016232967376709, "lr": 1.2860660487922616e-07, "epoch": 1.8561584840654608, "percentage": 92.81, "elapsed_time": "8:14:19", "remaining_time": "0:38:18", "throughput": 7404.72, "total_tokens": 219617128}
+{"current_steps": 2160, "total_steps": 2322, "loss": 0.3069549560546875, "lr": 1.2109595939289676e-07, "epoch": 1.8604651162790697, "percentage": 93.02, "elapsed_time": "8:15:29", "remaining_time": "0:37:09", "throughput": 7404.08, "total_tokens": 220118992}
+{"current_steps": 2165, "total_steps": 2322, "loss": 0.29118402004241944, "lr": 1.1380858677065177e-07, "epoch": 1.8647717484926787, "percentage": 93.24, "elapsed_time": "8:16:40", "remaining_time": "0:36:01", "throughput": 7403.13, "total_tokens": 220620384}
+{"current_steps": 2170, "total_steps": 2322, "loss": 0.3211177349090576, "lr": 1.0674482050389457e-07, "epoch": 1.8690783807062878, "percentage": 93.45, "elapsed_time": "8:17:47", "remaining_time": "0:34:52", "throughput": 7402.94, "total_tokens": 221104624}
+{"current_steps": 2175, "total_steps": 2322, "loss": 0.2935049533843994, "lr": 9.990498385115066e-08, "epoch": 1.8733850129198966, "percentage": 93.67, "elapsed_time": "8:18:53", "remaining_time": "0:33:43", "throughput": 7405.32, "total_tokens": 221664440}
+{"current_steps": 2180, "total_steps": 2322, "loss": 0.2967803478240967, "lr": 9.328938982327396e-08, "epoch": 1.8776916451335057, "percentage": 93.88, "elapsed_time": "8:19:54", "remaining_time": "0:32:33", "throughput": 7406.82, "total_tokens": 222166280}
+{"current_steps": 2185, "total_steps": 2322, "loss": 0.3003566265106201, "lr": 8.68983411691221e-08, "epoch": 1.8819982773471144, "percentage": 94.1, "elapsed_time": "8:20:54", "remaining_time": "0:31:24", "throughput": 7408.73, "total_tokens": 222663120}
+{"current_steps": 2190, "total_steps": 2322, "loss": 0.30220422744750974, "lr": 8.073213036170258e-08, "epoch": 1.8863049095607236, "percentage": 94.32, "elapsed_time": "8:21:57", "remaining_time": "0:30:15", "throughput": 7410.09, "total_tokens": 223174608}
+{"current_steps": 2195, "total_steps": 2322, "loss": 0.3004784107208252, "lr": 7.479103958478783e-08, "epoch": 1.8906115417743323, "percentage": 94.53, "elapsed_time": "8:22:58", "remaining_time": "0:29:06", "throughput": 7413.45, "total_tokens": 223724416}
+{"current_steps": 2200, "total_steps": 2322, "loss": 0.30257329940795896, "lr": 6.907534072000177e-08, "epoch": 1.8949181739879415, "percentage": 94.75, "elapsed_time": "8:24:17", "remaining_time": "0:27:57", "throughput": 7411.52, "total_tokens": 224257552}
+{"current_steps": 2205, "total_steps": 2322, "loss": 0.2995903491973877, "lr": 6.358529533437796e-08, "epoch": 1.8992248062015504, "percentage": 94.96, "elapsed_time": "8:25:31", "remaining_time": "0:26:49", "throughput": 7410.79, "total_tokens": 224783568}
+{"current_steps": 2210, "total_steps": 2322, "loss": 0.29779205322265623, "lr": 5.832115466838872e-08, "epoch": 1.9035314384151594, "percentage": 95.18, "elapsed_time": "8:26:47", "remaining_time": "0:25:41", "throughput": 7408.74, "total_tokens": 225281672}
+{"current_steps": 2215, "total_steps": 2322, "loss": 0.3170241117477417, "lr": 5.3283159624448745e-08, "epoch": 1.9078380706287683, "percentage": 95.39, "elapsed_time": "8:27:46", "remaining_time": "0:24:31", "throughput": 7411.96, "total_tokens": 225814656}
+{"current_steps": 2220, "total_steps": 2322, "loss": 0.3137502670288086, "lr": 4.847154075589e-08, "epoch": 1.9121447028423773, "percentage": 95.61, "elapsed_time": "8:28:53", "remaining_time": "0:23:22", "throughput": 7412.43, "total_tokens": 226330440}
+{"current_steps": 2225, "total_steps": 2322, "loss": 0.3071382999420166, "lr": 4.3886518256411325e-08, "epoch": 1.9164513350559862, "percentage": 95.82, "elapsed_time": "8:30:05", "remaining_time": "0:22:14", "throughput": 7411.64, "total_tokens": 226837032}
+{"current_steps": 2230, "total_steps": 2322, "loss": 0.30357215404510496, "lr": 3.9528301950000345e-08, "epoch": 1.9207579672695951, "percentage": 96.04, "elapsed_time": "8:31:21", "remaining_time": "0:21:05", "throughput": 7409.97, "total_tokens": 227345616}
+{"current_steps": 2235, "total_steps": 2322, "loss": 0.3153378963470459, "lr": 3.539709128133395e-08, "epoch": 1.9250645994832043, "percentage": 96.25, "elapsed_time": "8:32:31", "remaining_time": "0:19:57", "throughput": 7409.19, "total_tokens": 227840184}
+{"current_steps": 2240, "total_steps": 2322, "loss": 0.30716452598571775, "lr": 3.149307530664891e-08, "epoch": 1.929371231696813, "percentage": 96.47, "elapsed_time": "8:33:37", "remaining_time": "0:18:48", "throughput": 7409.83, "total_tokens": 228355216}
+{"current_steps": 2245, "total_steps": 2322, "loss": 0.2936999320983887, "lr": 2.7816432685091598e-08, "epoch": 1.9336778639104222, "percentage": 96.68, "elapsed_time": "8:34:42", "remaining_time": "0:17:39", "throughput": 7411.34, "total_tokens": 228882600}
+{"current_steps": 2250, "total_steps": 2322, "loss": 0.3067632675170898, "lr": 2.4367331670539506e-08, "epoch": 1.937984496124031, "percentage": 96.9, "elapsed_time": "8:35:54", "remaining_time": "0:16:30", "throughput": 7410.79, "total_tokens": 229400664}
+{"current_steps": 2255, "total_steps": 2322, "loss": 0.3079858303070068, "lr": 2.1145930103904645e-08, "epoch": 1.94229112833764, "percentage": 97.11, "elapsed_time": "8:37:02", "remaining_time": "0:15:21", "throughput": 7411.18, "total_tokens": 229912008}
+{"current_steps": 2260, "total_steps": 2322, "loss": 0.3009981632232666, "lr": 1.8152375405909305e-08, "epoch": 1.9465977605512488, "percentage": 97.33, "elapsed_time": "8:38:18", "remaining_time": "0:14:13", "throughput": 7410.2, "total_tokens": 230446552}
+{"current_steps": 2265, "total_steps": 2322, "loss": 0.3050046920776367, "lr": 1.538680457033814e-08, "epoch": 1.950904392764858, "percentage": 97.55, "elapsed_time": "8:39:33", "remaining_time": "0:13:04", "throughput": 7409.39, "total_tokens": 230974000}
+{"current_steps": 2270, "total_steps": 2322, "loss": 0.3118910789489746, "lr": 1.284934415776984e-08, "epoch": 1.9552110249784669, "percentage": 97.76, "elapsed_time": "8:40:49", "remaining_time": "0:11:55", "throughput": 7408.48, "total_tokens": 231510712}
+{"current_steps": 2275, "total_steps": 2322, "loss": 0.28809175491333006, "lr": 1.0540110289786742e-08, "epoch": 1.9595176571920758, "percentage": 97.98, "elapsed_time": "8:41:54", "remaining_time": "0:10:46", "throughput": 7407.35, "total_tokens": 231960472}
+{"current_steps": 2280, "total_steps": 2322, "loss": 0.3013895511627197, "lr": 8.459208643659122e-09, "epoch": 1.9638242894056848, "percentage": 98.19, "elapsed_time": "8:43:10", "remaining_time": "0:09:38", "throughput": 7405.56, "total_tokens": 232466016}
+{"current_steps": 2285, "total_steps": 2322, "loss": 0.2956396579742432, "lr": 6.606734447507923e-09, "epoch": 1.9681309216192937, "percentage": 98.41, "elapsed_time": "8:44:15", "remaining_time": "0:08:29", "throughput": 7407.36, "total_tokens": 233001544}
+{"current_steps": 2290, "total_steps": 2322, "loss": 0.3004451751708984, "lr": 4.982772475951026e-09, "epoch": 1.9724375538329026, "percentage": 98.62, "elapsed_time": "8:45:24", "remaining_time": "0:07:20", "throughput": 7406.43, "total_tokens": 233486880}
+{"current_steps": 2295, "total_steps": 2322, "loss": 0.3195258617401123, "lr": 3.587397046219132e-09, "epoch": 1.9767441860465116, "percentage": 98.84, "elapsed_time": "8:46:32", "remaining_time": "0:06:11", "throughput": 7406.71, "total_tokens": 233992944}
+{"current_steps": 2300, "total_steps": 2322, "loss": 0.3012960433959961, "lr": 2.4206720147573748e-09, "epoch": 1.9810508182601207, "percentage": 99.05, "elapsed_time": "8:47:42", "remaining_time": "0:05:02", "throughput": 7406.53, "total_tokens": 234505824}
+{"current_steps": 2305, "total_steps": 2322, "loss": 0.2931736707687378, "lr": 1.4826507743032071e-09, "epoch": 1.9853574504737295, "percentage": 99.27, "elapsed_time": "8:48:47", "remaining_time": "0:03:54", "throughput": 7406.43, "total_tokens": 234988896}
+{"current_steps": 2310, "total_steps": 2322, "loss": 0.3052358865737915, "lr": 7.733762514411381e-10, "epoch": 1.9896640826873386, "percentage": 99.48, "elapsed_time": "8:50:02", "remaining_time": "0:02:45", "throughput": 7404.66, "total_tokens": 235490216}
+{"current_steps": 2315, "total_steps": 2322, "loss": 0.29867162704467776, "lr": 2.928809046398584e-10, "epoch": 1.9939707149009473, "percentage": 99.7, "elapsed_time": "8:51:19", "remaining_time": "0:01:36", "throughput": 7403.26, "total_tokens": 236010080}
+{"current_steps": 2320, "total_steps": 2322, "loss": 0.3062272548675537, "lr": 4.118672276620661e-11, "epoch": 1.9982773471145565, "percentage": 99.91, "elapsed_time": "8:52:38", "remaining_time": "0:00:27", "throughput": 7401.66, "total_tokens": 236545192}
+{"current_steps": 2322, "total_steps": 2322, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "8:53:23", "remaining_time": "0:00:00", "throughput": 7397.73, "total_tokens": 236749552}
--- a/trainer_state.json
+++ b/trainer_state.json
--- a/training_args.bin
+++ b/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e1e995bcc45bb5d176ef43a23610ceefbca95a9495933ad7c576c2d32115034
+size 7697
--- a/training_loss.png
+++ b/training_loss.png