初始化项目，由ModelHub XC社区提供模型

Model: lldois/v26_v21_product_rec_rebalance_lr22e6_ep032 Source: Original Platform
2026-08-19 05:59:17 +08:00
commit c3bdd76b11
17 changed files with 2831 additions and 0 deletions
--- a/.gitattributes
+++ b/.gitattributes
@@ -0,0 +1,36 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
--- a/README.md
+++ b/README.md
@@ -0,0 +1,58 @@
+---
+library_name: transformers
+license: other
+base_model: lldois/v21_v15_user_repair_rec_fast_lr4e6_ep055
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: v26_v21_product_rec_rebalance_lr22e6_ep032
+  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. -->
+
+# v26_v21_product_rec_rebalance_lr22e6_ep032
+
+This model is a fine-tuned version of [lldois/v21_v15_user_repair_rec_fast_lr4e6_ep055](https://huggingface.co/lldois/v21_v15_user_repair_rec_fast_lr4e6_ep055) on the exp_v21_product_rec_rebalance dataset.
+
+## 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: 2.2e-06
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 202607261
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 4
+- 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
+- lr_scheduler_warmup_steps: 0.02
+- num_epochs: 0.32
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.6.0
+- Pytorch 2.7.1+cu126
+- Datasets 4.0.0
+- Tokenizers 0.22.2
--- a/all_results.json
+++ b/all_results.json
@@ -0,0 +1,8 @@
+{
+    "epoch": 0.3204837490551776,
+    "total_flos": 5.064451518240154e+16,
+    "train_loss": 0.745630294084549,
+    "train_runtime": 819.7494,
+    "train_samples_per_second": 0.516,
+    "train_steps_per_second": 0.129
+}
--- 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/config.json
+++ b/config.json
@@ -0,0 +1,63 @@
+{
+  "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": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "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": 176253
+}
--- a/experiment_recipe.json
+++ b/experiment_recipe.json
@@ -0,0 +1,62 @@
+{
+  "created_at": "2026-07-10 23:45:28",
+  "run": {
+    "name": "v26_v21_product_rec_rebalance_lr22e6_ep032",
+    "dataset": "v21_product_rec_rebalance",
+    "model_path": "/home/ll/llm4rec/experiments/outputs/v21_v15_user_repair_rec_fast_lr4e6_ep055",
+    "lr": "2.2e-6",
+    "epochs": 0.32,
+    "warmup": 0.02,
+    "scheduler": "cosine",
+    "seed": 202607261,
+    "note": "v21 continuation. Rec rebalance from the best user1 checkpoint: heavier product/video rec route supervision, item replay, and only light user replay. Goal: preserve user1 while raising rec1/rec2.",
+    "config": "/home/ll/llm4rec/experiments/configs/v26_v21_product_rec_rebalance_lr22e6_ep032.yaml"
+  },
+  "dataset_manifest": {
+    "name": "v21_product_rec_rebalance",
+    "path": "/home/ll/llm4rec/experiments/data/v21_product_rec_rebalance.jsonl",
+    "records": 37594,
+    "groups": {
+      "rec": 24000,
+      "item": 11194,
+      "user": 2400
+    },
+    "variants": {
+      "rec_short_think_final": 9600,
+      "item_no_think_direct_final": 2805,
+      "rec_no_think_direct_final": 14400,
+      "user_strict_logic": 904,
+      "user_strict_array": 1094,
+      "item_short_think": 2792,
+      "item_compact_cot": 5597,
+      "user_extra_no_think_logic": 402
+    },
+    "sha256": "5b58f82b4dfaed0a3bee1dc5ae24ee4406ae40c70038e6bca271ca61b8fe50fa",
+    "seed": 202607112
+  },
+  "dataset_recipe": "v21 continuation: heavier product/video rec repair, full item replay, and light user replay. Tests whether highest-user checkpoint can recover recommendation metrics.",
+  "cot_policy": "Preserve /think reasoning supervision and do not use v7_final_only as a CoT training base. For /no_think prompts, train pure final answers without generated <think> tags. This is route-specific behavior, not global CoT removal.",
+  "raw_counts": {
+    "rec": 19204,
+    "item": 10384,
+    "user": 2892
+  },
+  "eval_observations": {
+    "v07": "best local score so far: total=0.8978, eval_time≈47.3min; fast final outputs likely help.",
+    "v15": "best CoT-preserving score so far: total=0.8778, eval_time≈70.1min; logs show repeated tokens, JSON shell errors, prompt leakage, and verbose /no_think outputs.",
+    "v19": "best CoT-native continuation so far: total=0.8855, eval_time≈48.1min; user1 and rec4 improved but world dropped.",
+    "v20": "v7 final-only continuation with CoT restore failed as a CoT route: total=0.8527, item fell to 0.1840; do not use v7 as future CoT base.",
+    "v22": "scratch official-base 3 epoch clean CoT underperformed: total=0.8217; item/world preserved but user and rec2 are weak.",
+    "v23": "scratch official-base 5 epoch low-LR guard failed badly: total=0.6990; item/user collapse suggests long scratch SFT is not viable with current data mix.",
+    "v24": "v15 light repair is best among v22-v24 but still only total=0.8364; user2 improves but rec/world do not recover.",
+    "v16": "CoT pattern rewrite failed: total=0.7912; logs show malformed user JSON and fragmented recommendation reasoning.",
+    "v18": "low-LR mixed replay from v12 failed: total=0.8340; item/world dropped and rec outputs mixed text/itemic/think tags."
+  },
+  "script": "/home/ll/llm4rec/experiments/run_experiments.py",
+  "script_sha256": "f752ec011aae9158402c880ebe487650272242b5b35d45bd4a6d9ec1c7dc33d5",
+  "deadline": "2026-07-11 09:00:00 +0800",
+  "reproduce": {
+    "prepare_command": "EXPERIMENT_PREPARE_ONLY=1 python3 experiments/run_experiments.py",
+    "train_command": "CUDA_VISIBLE_DEVICES=<gpu> bash -lc 'source /home/ll/llm4rec/demo/LLaMA-Factory/.venv/bin/activate && llamafactory-cli train /home/ll/llm4rec/experiments/configs/v26_v21_product_rec_rebalance_lr22e6_ep032.yaml'"
+  }
+}
--- 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:d62b80611c55e38f996606a87abf1760fd7dd44aceebbd130c2f4023b374c657
+size 1602902832
--- a/runs/Jul10_23-45-59_ubuntu-System-Product-Name/events.out.tfevents.1783698359.ubuntu-System-Product-Name.3126867.0
+++ b/runs/Jul10_23-45-59_ubuntu-System-Product-Name/events.out.tfevents.1783698359.ubuntu-System-Product-Name.3126867.0
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce2c6dc40a6d1425d8bf419c655fbcb25f2e99ecdc25476909a09b19649b6003
+size 10665
--- a/tokenizer.json
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd4d15f596979aecbc11ba12668cb21c9fb8452ce1235cd9d7878cc950170421
+size 16016554
--- a/tokenizer_config.json
+++ b/tokenizer_config.json
@@ -0,0 +1,16 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "is_local": true,
+  "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.log
+++ b/train.log
--- a/train_results.json
+++ b/train_results.json
@@ -0,0 +1,8 @@
+{
+    "epoch": 0.3204837490551776,
+    "total_flos": 5.064451518240154e+16,
+    "train_loss": 0.745630294084549,
+    "train_runtime": 819.7494,
+    "train_samples_per_second": 0.516,
+    "train_steps_per_second": 0.129
+}
--- a/trainer_log.jsonl
+++ b/trainer_log.jsonl
@@ -0,0 +1,22 @@
+{"current_steps": 5, "total_steps": 106, "loss": 0.9165916442871094, "lr": 2.199488371985687e-06, "epoch": 0.015117157974300832, "percentage": 4.72, "elapsed_time": "0:00:35", "remaining_time": "0:12:05"}
+{"current_steps": 10, "total_steps": 106, "loss": 0.8236251831054687, "lr": 2.1816313149006775e-06, "epoch": 0.030234315948601664, "percentage": 9.43, "elapsed_time": "0:01:14", "remaining_time": "0:11:56"}
+{"current_steps": 15, "total_steps": 106, "loss": 0.8143067359924316, "lr": 2.1386668040096687e-06, "epoch": 0.045351473922902494, "percentage": 14.15, "elapsed_time": "0:01:51", "remaining_time": "0:11:16"}
+{"current_steps": 20, "total_steps": 106, "loss": 0.6190191268920898, "lr": 2.071592156473907e-06, "epoch": 0.06046863189720333, "percentage": 18.87, "elapsed_time": "0:02:30", "remaining_time": "0:10:45"}
+{"current_steps": 25, "total_steps": 106, "loss": 0.7870875835418701, "lr": 1.981964347938752e-06, "epoch": 0.07558578987150416, "percentage": 23.58, "elapsed_time": "0:03:09", "remaining_time": "0:10:14"}
+{"current_steps": 30, "total_steps": 106, "loss": 0.6626455307006835, "lr": 1.8718638711114028e-06, "epoch": 0.09070294784580499, "percentage": 28.3, "elapsed_time": "0:03:48", "remaining_time": "0:09:38"}
+{"current_steps": 35, "total_steps": 106, "loss": 0.6299720287322998, "lr": 1.7438464421567664e-06, "epoch": 0.10582010582010581, "percentage": 33.02, "elapsed_time": "0:04:28", "remaining_time": "0:09:03"}
+{"current_steps": 40, "total_steps": 106, "loss": 0.6886922359466553, "lr": 1.6008836759306283e-06, "epoch": 0.12093726379440665, "percentage": 37.74, "elapsed_time": "0:05:07", "remaining_time": "0:08:26"}
+{"current_steps": 45, "total_steps": 106, "loss": 0.7747112274169922, "lr": 1.4462941071310436e-06, "epoch": 0.1360544217687075, "percentage": 42.45, "elapsed_time": "0:05:45", "remaining_time": "0:07:48"}
+{"current_steps": 50, "total_steps": 106, "loss": 0.6844994068145752, "lr": 1.2836661585450144e-06, "epoch": 0.15117157974300832, "percentage": 47.17, "elapsed_time": "0:06:24", "remaining_time": "0:07:10"}
+{"current_steps": 55, "total_steps": 106, "loss": 0.735169792175293, "lr": 1.1167748444962186e-06, "epoch": 0.16628873771730915, "percentage": 51.89, "elapsed_time": "0:07:04", "remaining_time": "0:06:33"}
+{"current_steps": 60, "total_steps": 106, "loss": 0.7788573265075683, "lr": 9.49494143021699e-07, "epoch": 0.18140589569160998, "percentage": 56.6, "elapsed_time": "0:07:43", "remaining_time": "0:05:55"}
+{"current_steps": 65, "total_steps": 106, "loss": 0.7242467880249024, "lr": 7.857070708454049e-07, "epoch": 0.1965230536659108, "percentage": 61.32, "elapsed_time": "0:08:23", "remaining_time": "0:05:17"}
+{"current_steps": 70, "total_steps": 106, "loss": 0.8203678131103516, "lr": 6.292155485404879e-07, "epoch": 0.21164021164021163, "percentage": 66.04, "elapsed_time": "0:08:59", "remaining_time": "0:04:37"}
+{"current_steps": 75, "total_steps": 106, "loss": 0.7025283336639404, "lr": 4.836521481425153e-07, "epoch": 0.22675736961451248, "percentage": 70.75, "elapsed_time": "0:09:37", "remaining_time": "0:03:58"}
+{"current_steps": 80, "total_steps": 106, "loss": 0.6882677555084229, "lr": 3.5239577177921575e-07, "epoch": 0.2418745275888133, "percentage": 75.47, "elapsed_time": "0:10:15", "remaining_time": "0:03:19"}
+{"current_steps": 85, "total_steps": 106, "loss": 0.7953117370605469, "lr": 2.3849321863334733e-07, "epoch": 0.25699168556311414, "percentage": 80.19, "elapsed_time": "0:10:55", "remaining_time": "0:02:42"}
+{"current_steps": 90, "total_steps": 106, "loss": 0.7164456844329834, "lr": 1.4458846087185652e-07, "epoch": 0.272108843537415, "percentage": 84.91, "elapsed_time": "0:11:32", "remaining_time": "0:02:03"}
+{"current_steps": 95, "total_steps": 106, "loss": 0.6571497917175293, "lr": 7.286127022951101e-08, "epoch": 0.2872260015117158, "percentage": 89.62, "elapsed_time": "0:12:10", "remaining_time": "0:01:24"}
+{"current_steps": 100, "total_steps": 106, "loss": 0.8477729797363281, "lr": 2.497661988236255e-08, "epoch": 0.30234315948601664, "percentage": 94.34, "elapsed_time": "0:12:50", "remaining_time": "0:00:46"}
+{"current_steps": 105, "total_steps": 106, "loss": 0.8160898208618164, "lr": 2.0460361241167792e-09, "epoch": 0.31746031746031744, "percentage": 99.06, "elapsed_time": "0:13:31", "remaining_time": "0:00:07"}
+{"current_steps": 106, "total_steps": 106, "epoch": 0.3204837490551776, "percentage": 100.0, "elapsed_time": "0:13:39", "remaining_time": "0:00:00"}
--- a/trainer_state.json
+++ b/trainer_state.json
@@ -0,0 +1,190 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3204837490551776,
+  "eval_steps": 500,
+  "global_step": 106,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.015117157974300832,
+      "grad_norm": 38.25,
+      "learning_rate": 2.199488371985687e-06,
+      "loss": 0.9165916442871094,
+      "step": 5
+    },
+    {
+      "epoch": 0.030234315948601664,
+      "grad_norm": 40.5,
+      "learning_rate": 2.1816313149006775e-06,
+      "loss": 0.8236251831054687,
+      "step": 10
+    },
+    {
+      "epoch": 0.045351473922902494,
+      "grad_norm": 40.25,
+      "learning_rate": 2.1386668040096687e-06,
+      "loss": 0.8143067359924316,
+      "step": 15
+    },
+    {
+      "epoch": 0.06046863189720333,
+      "grad_norm": 4.90625,
+      "learning_rate": 2.071592156473907e-06,
+      "loss": 0.6190191268920898,
+      "step": 20
+    },
+    {
+      "epoch": 0.07558578987150416,
+      "grad_norm": 38.75,
+      "learning_rate": 1.981964347938752e-06,
+      "loss": 0.7870875835418701,
+      "step": 25
+    },
+    {
+      "epoch": 0.09070294784580499,
+      "grad_norm": 4.625,
+      "learning_rate": 1.8718638711114028e-06,
+      "loss": 0.6626455307006835,
+      "step": 30
+    },
+    {
+      "epoch": 0.10582010582010581,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.7438464421567664e-06,
+      "loss": 0.6299720287322998,
+      "step": 35
+    },
+    {
+      "epoch": 0.12093726379440665,
+      "grad_norm": 4.375,
+      "learning_rate": 1.6008836759306283e-06,
+      "loss": 0.6886922359466553,
+      "step": 40
+    },
+    {
+      "epoch": 0.1360544217687075,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.4462941071310436e-06,
+      "loss": 0.7747112274169922,
+      "step": 45
+    },
+    {
+      "epoch": 0.15117157974300832,
+      "grad_norm": 4.5,
+      "learning_rate": 1.2836661585450144e-06,
+      "loss": 0.6844994068145752,
+      "step": 50
+    },
+    {
+      "epoch": 0.16628873771730915,
+      "grad_norm": 36.75,
+      "learning_rate": 1.1167748444962186e-06,
+      "loss": 0.735169792175293,
+      "step": 55
+    },
+    {
+      "epoch": 0.18140589569160998,
+      "grad_norm": 30.375,
+      "learning_rate": 9.49494143021699e-07,
+      "loss": 0.7788573265075683,
+      "step": 60
+    },
+    {
+      "epoch": 0.1965230536659108,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.857070708454049e-07,
+      "loss": 0.7242467880249024,
+      "step": 65
+    },
+    {
+      "epoch": 0.21164021164021163,
+      "grad_norm": 31.875,
+      "learning_rate": 6.292155485404879e-07,
+      "loss": 0.8203678131103516,
+      "step": 70
+    },
+    {
+      "epoch": 0.22675736961451248,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.836521481425153e-07,
+      "loss": 0.7025283336639404,
+      "step": 75
+    },
+    {
+      "epoch": 0.2418745275888133,
+      "grad_norm": 4.90625,
+      "learning_rate": 3.5239577177921575e-07,
+      "loss": 0.6882677555084229,
+      "step": 80
+    },
+    {
+      "epoch": 0.25699168556311414,
+      "grad_norm": 5.375,
+      "learning_rate": 2.3849321863334733e-07,
+      "loss": 0.7953117370605469,
+      "step": 85
+    },
+    {
+      "epoch": 0.272108843537415,
+      "grad_norm": 14.5,
+      "learning_rate": 1.4458846087185652e-07,
+      "loss": 0.7164456844329834,
+      "step": 90
+    },
+    {
+      "epoch": 0.2872260015117158,
+      "grad_norm": 4.78125,
+      "learning_rate": 7.286127022951101e-08,
+      "loss": 0.6571497917175293,
+      "step": 95
+    },
+    {
+      "epoch": 0.30234315948601664,
+      "grad_norm": 29.875,
+      "learning_rate": 2.497661988236255e-08,
+      "loss": 0.8477729797363281,
+      "step": 100
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 5.46875,
+      "learning_rate": 2.0460361241167792e-09,
+      "loss": 0.8160898208618164,
+      "step": 105
+    },
+    {
+      "epoch": 0.3204837490551776,
+      "step": 106,
+      "total_flos": 5.064451518240154e+16,
+      "train_loss": 0.745630294084549,
+      "train_runtime": 819.7494,
+      "train_samples_per_second": 0.516,
+      "train_steps_per_second": 0.129
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 106,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.064451518240154e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
--- a/training_args.bin
+++ b/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef46209b41f39d4aa2e67bc28685960d530fd312c6d3b2d015112ed2236ced63
+size 6033
--- a/training_loss.png
+++ b/training_loss.png