first commit

README.md

@@ -1,3 +1,81 @@
----
-license: apache-2.0
----
+---
+base_model: /mnt/cache/luzimu/rlhf_math/alignment-handbook/outs/Mistral-7B-v0.1-lce
+tags:
+- alignment-handbook
+- generated_from_trainer
+datasets:
+- /mnt/cache/luzimu/rlhf_math/data/controled_steps_math_gsm8k_lce_dpo_ascend_lim2_lim3_add_dpo1x1
+model-index:
+- name: Mistral-7B-v0.1-lce_controled_steps_dpo_ascend_lim2_lim3_add_dpo1x1
+  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. -->
+
+# Mistral-7B-v0.1-lce_controled_steps_dpo_ascend_lim2_lim3_add_dpo1x1
+
+This model is a fine-tuned version of [/mnt/cache/luzimu/rlhf_math/alignment-handbook/outs/Mistral-7B-v0.1-lce](https://huggingface.co//mnt/cache/luzimu/rlhf_math/alignment-handbook/outs/Mistral-7B-v0.1-lce) on the /mnt/cache/luzimu/rlhf_math/data/controled_steps_math_gsm8k_lce_dpo_ascend_lim2_lim3_add_dpo1x1 dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.1793
+- Rewards/chosen: 0.2587
+- Rewards/rejected: -7.0301
+- Rewards/accuracies: 0.8947
+- Rewards/margins: 7.2889
+- Logps/rejected: -253.7773
+- Logps/chosen: -80.3105
+- Logits/rejected: -2.3417
+- Logits/chosen: -2.3846
+
+## 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-07
+- train_batch_size: 2
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 64
+- total_eval_batch_size: 16
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 2
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
+|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
+| 0.3963        | 0.21  | 100  | 0.3636          | 1.8634         | -0.1518          | 0.8816             | 2.0152          | -184.9944      | -64.2644     | -2.7112         | -2.7505       |
+| 0.2849        | 0.43  | 200  | 0.2598          | 0.7706         | -3.7221          | 0.8816             | 4.4927          | -220.6974      | -75.1921     | -2.5067         | -2.5475       |
+| 0.2496        | 0.64  | 300  | 0.2295          | 0.9323         | -4.2717          | 0.8684             | 5.2040          | -226.1934      | -73.5753     | -2.5080         | -2.5494       |
+| 0.2331        | 0.86  | 400  | 0.2089          | 0.7871         | -4.8912          | 0.8684             | 5.6783          | -232.3884      | -75.0269     | -2.4967         | -2.5382       |
+| 0.0874        | 1.07  | 500  | 0.1872          | 0.6345         | -5.7444          | 0.8816             | 6.3789          | -240.9202      | -76.5527     | -2.4323         | -2.4761       |
+| 0.1217        | 1.28  | 600  | 0.1832          | 0.2282         | -6.6907          | 0.8684             | 6.9188          | -250.3827      | -80.6161     | -2.3741         | -2.4172       |
+| 0.0966        | 1.5   | 700  | 0.1807          | 0.1849         | -7.0125          | 0.8816             | 7.1975          | -253.6012      | -81.0485     | -2.3503         | -2.3940       |
+| 0.0755        | 1.71  | 800  | 0.1802          | 0.3224         | -6.9539          | 0.8947             | 7.2763          | -253.0150      | -79.6739     | -2.3437         | -2.3867       |
+| 0.1177        | 1.93  | 900  | 0.1793          | 0.2587         | -7.0301          | 0.8947             | 7.2889          | -253.7773      | -80.3105     | -2.3417         | -2.3846       |
+
+
+### Framework versions
+
+- Transformers 4.38.2
+- Pytorch 2.1.2
+- Datasets 2.14.6
+- Tokenizers 0.15.2

added_tokens.json

@@ -0,0 +1,10 @@
+{
+  "<|assistant|>": 32002,
+  "<|code|>": 32004,
+  "<|endofblock|>": 32006,
+  "<|endofmessage|>": 32007,
+  "<|execution|>": 32005,
+  "<|system|>": 32001,
+  "<|text|>": 32003,
+  "<|user|>": 32000
+}

all_results.json

@@ -0,0 +1,21 @@
+{
+    "epoch": 2.0,
+    "eval_logits/chosen": -2.3846213817596436,
+    "eval_logits/rejected": -2.3414511680603027,
+    "eval_logps/chosen": -80.37240600585938,
+    "eval_logps/rejected": -253.9702911376953,
+    "eval_loss": 0.17943651974201202,
+    "eval_rewards/accuracies": 0.8947368264198303,
+    "eval_rewards/chosen": 0.25255322456359863,
+    "eval_rewards/margins": 7.301982879638672,
+    "eval_rewards/rejected": -7.049429416656494,
+    "eval_runtime": 27.9087,
+    "eval_samples": 301,
+    "eval_samples_per_second": 10.785,
+    "eval_steps_per_second": 0.681,
+    "train_loss": 0.21607095603003493,
+    "train_runtime": 14362.1771,
+    "train_samples": 29893,
+    "train_samples_per_second": 4.163,
+    "train_steps_per_second": 0.065
+}

config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "/mnt/cache/luzimu/rlhf_math/alignment-handbook/outs/Mistral-7B-v0.1-lce",
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000.0,
+  "sliding_window": 4096,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.38.2",
+  "use_cache": true,
+  "vocab_size": 32128
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.0,
+    "eval_logits/chosen": -2.3846213817596436,
+    "eval_logits/rejected": -2.3414511680603027,
+    "eval_logps/chosen": -80.37240600585938,
+    "eval_logps/rejected": -253.9702911376953,
+    "eval_loss": 0.17943651974201202,
+    "eval_rewards/accuracies": 0.8947368264198303,
+    "eval_rewards/chosen": 0.25255322456359863,
+    "eval_rewards/margins": 7.301982879638672,
+    "eval_rewards/rejected": -7.049429416656494,
+    "eval_runtime": 27.9087,
+    "eval_samples": 301,
+    "eval_samples_per_second": 10.785,
+    "eval_steps_per_second": 0.681
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.38.2"
+}

model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:069598f45d8fd3590f09b131ccc53f5e465fcad0e0fa0311e573059bc11981dc
+size 4944210912

model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41642eb1352d6e8c6223d417778bbe985f677e0528884a6847c79e32de850e97
+size 4999819336

model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a65eeb14cb3c2bd4bd2460f5f56989d6d02df8b51446409f430afcfb4b042bda
+size 4541564920

model.safetensors.index.json

@@ -0,0 +1,298 @@
+{
+  "metadata": {
+    "total_size": 14485561344
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00003-of-00003.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.norm.weight": "model-00003-of-00003.safetensors"
+  }
+}

runs/Apr13_21-08-49_pt-4rcsf1vl-worker-0/events.out.tfevents.1713013930.pt-4rcsf1vl-worker-0.38.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3889c86e1e64be477579961cb4a6af53bfa372d24eec5019ffff3f30a33a9f0
+size 76575

runs/Apr13_21-08-49_pt-4rcsf1vl-worker-0/events.out.tfevents.1713028337.pt-4rcsf1vl-worker-0.38.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c1c008b168faa818b02fe2fed64b827c767cf28e925a7a453187472ff8bb45d
+size 828

special_tokens_map.json

@@ -0,0 +1,40 @@
+{
+  "additional_special_tokens": [
+    "<|user|>",
+    "<|system|>",
+    "<|assistant|>",
+    "<|text|>",
+    "<|code|>",
+    "<|execution|>",
+    "<|endofblock|>",
+    "<|endofmessage|>"
+  ],
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

tokenizer_config.json

@@ -0,0 +1,116 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32000": {
+      "content": "<|user|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32001": {
+      "content": "<|system|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32002": {
+      "content": "<|assistant|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32003": {
+      "content": "<|text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32004": {
+      "content": "<|code|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32005": {
+      "content": "<|execution|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32006": {
+      "content": "<|endofblock|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32007": {
+      "content": "<|endofmessage|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|user|>",
+    "<|system|>",
+    "<|assistant|>",
+    "<|text|>",
+    "<|code|>",
+    "<|execution|>",
+    "<|endofblock|>",
+    "<|endofmessage|>"
+  ],
+  "bos_token": "<s>",
+  "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>' }}{% elif message['role'] == 'system' %}\n{{ '<|system|>' }}{% elif message['role'] == 'assistant_correct' %}\n{{ '<|assistant|>' }}{% endif %}\n{% for block in message['content'] %}\n{% if block['type'] == 'text' %}\n{{ '<|text|>' }}{% elif block['type'] == 'code' %}\n{{ '<|code|>' }}{% elif block['type'] == 'execution' %}\n{{ '<|execution|>' }}{% endif %}\n{{ block['content'] + '<|endofblock|>' }}{% endfor %}\n{% if message['role'] != 'assistant_correct' %}\n{{ '<|endofmessage|>' }}{% endif %}{% endfor %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 2048,
+  "pad_token": "</s>",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": true
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "train_loss": 0.21607095603003493,
+    "train_runtime": 14362.1771,
+    "train_samples": 29893,
+    "train_samples_per_second": 4.163,
+    "train_steps_per_second": 0.065
+}

trainer_state.json

@@ -0,0 +1,1584 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9994648113459994,
+  "eval_steps": 100,
+  "global_step": 934,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 233.47004915999355,
+      "learning_rate": 1.0638297872340424e-09,
+      "logits/chosen": -2.9313061237335205,
+      "logits/rejected": -2.9551453590393066,
+      "logps/chosen": -84.83019256591797,
+      "logps/rejected": -122.72698974609375,
+      "loss": 0.6931,
+      "rewards/accuracies": 0.0,
+      "rewards/chosen": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/rejected": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 260.5751075825045,
+      "learning_rate": 1.0638297872340425e-08,
+      "logits/chosen": -2.9027864933013916,
+      "logits/rejected": -2.9143683910369873,
+      "logps/chosen": -89.16300964355469,
+      "logps/rejected": -185.06263732910156,
+      "loss": 0.6955,
+      "rewards/accuracies": 0.4236111044883728,
+      "rewards/chosen": 0.003953192848712206,
+      "rewards/margins": -6.558249151566997e-05,
+      "rewards/rejected": 0.004018775187432766,
+      "step": 10
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 208.3552637601932,
+      "learning_rate": 2.127659574468085e-08,
+      "logits/chosen": -2.913933038711548,
+      "logits/rejected": -2.928837537765503,
+      "logps/chosen": -80.42648315429688,
+      "logps/rejected": -182.06158447265625,
+      "loss": 0.6916,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.014185728505253792,
+      "rewards/margins": 0.009901548735797405,
+      "rewards/rejected": 0.004284179303795099,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 230.6593263920013,
+      "learning_rate": 3.191489361702128e-08,
+      "logits/chosen": -2.9638733863830566,
+      "logits/rejected": -2.947378396987915,
+      "logps/chosen": -98.21806335449219,
+      "logps/rejected": -209.5995635986328,
+      "loss": 0.6746,
+      "rewards/accuracies": 0.6187499761581421,
+      "rewards/chosen": 0.06607931107282639,
+      "rewards/margins": 0.030630504712462425,
+      "rewards/rejected": 0.03544881194829941,
+      "step": 30
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 229.5001403960732,
+      "learning_rate": 4.25531914893617e-08,
+      "logits/chosen": -2.8966455459594727,
+      "logits/rejected": -2.928231716156006,
+      "logps/chosen": -89.00160217285156,
+      "logps/rejected": -181.76254272460938,
+      "loss": 0.6351,
+      "rewards/accuracies": 0.737500011920929,
+      "rewards/chosen": 0.2816316783428192,
+      "rewards/margins": 0.13491405546665192,
+      "rewards/rejected": 0.1467176377773285,
+      "step": 40
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 156.2561754732514,
+      "learning_rate": 5.3191489361702123e-08,
+      "logits/chosen": -2.900092601776123,
+      "logits/rejected": -2.8932037353515625,
+      "logps/chosen": -83.15284729003906,
+      "logps/rejected": -171.3411102294922,
+      "loss": 0.5936,
+      "rewards/accuracies": 0.7437499761581421,
+      "rewards/chosen": 0.5719220638275146,
+      "rewards/margins": 0.29314228892326355,
+      "rewards/rejected": 0.2787797451019287,
+      "step": 50
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 132.07680968174566,
+      "learning_rate": 6.382978723404255e-08,
+      "logits/chosen": -2.891676425933838,
+      "logits/rejected": -2.9195587635040283,
+      "logps/chosen": -74.60004425048828,
+      "logps/rejected": -182.98202514648438,
+      "loss": 0.5168,
+      "rewards/accuracies": 0.824999988079071,
+      "rewards/chosen": 1.0686428546905518,
+      "rewards/margins": 0.6195236444473267,
+      "rewards/rejected": 0.4491192400455475,
+      "step": 60
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 151.5773318247949,
+      "learning_rate": 7.446808510638298e-08,
+      "logits/chosen": -2.9197444915771484,
+      "logits/rejected": -2.9284892082214355,
+      "logps/chosen": -74.910888671875,
+      "logps/rejected": -172.590087890625,
+      "loss": 0.5056,
+      "rewards/accuracies": 0.824999988079071,
+      "rewards/chosen": 1.4046766757965088,
+      "rewards/margins": 0.7043676376342773,
+      "rewards/rejected": 0.7003089189529419,
+      "step": 70
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 130.65989464011713,
+      "learning_rate": 8.51063829787234e-08,
+      "logits/chosen": -2.9318323135375977,
+      "logits/rejected": -2.897862195968628,
+      "logps/chosen": -67.76197052001953,
+      "logps/rejected": -189.8668670654297,
+      "loss": 0.4526,
+      "rewards/accuracies": 0.84375,
+      "rewards/chosen": 1.649476408958435,
+      "rewards/margins": 1.2739518880844116,
+      "rewards/rejected": 0.3755245506763458,
+      "step": 80
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 135.89857787297694,
+      "learning_rate": 9.574468085106384e-08,
+      "logits/chosen": -2.870614528656006,
+      "logits/rejected": -2.8664841651916504,
+      "logps/chosen": -72.87008666992188,
+      "logps/rejected": -173.2789764404297,
+      "loss": 0.4347,
+      "rewards/accuracies": 0.8125,
+      "rewards/chosen": 1.8745962381362915,
+      "rewards/margins": 1.2003813982009888,
+      "rewards/rejected": 0.674214780330658,
+      "step": 90
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 147.9607240979239,
+      "learning_rate": 9.998741174712532e-08,
+      "logits/chosen": -2.8332231044769287,
+      "logits/rejected": -2.846297264099121,
+      "logps/chosen": -74.4647445678711,
+      "logps/rejected": -185.64999389648438,
+      "loss": 0.3963,
+      "rewards/accuracies": 0.856249988079071,
+      "rewards/chosen": 1.7484337091445923,
+      "rewards/margins": 1.455592393875122,
+      "rewards/rejected": 0.29284122586250305,
+      "step": 100
+    },
+    {
+      "epoch": 0.21,
+      "eval_logits/chosen": -2.750455856323242,
+      "eval_logits/rejected": -2.711240530014038,
+      "eval_logps/chosen": -64.26441192626953,
+      "eval_logps/rejected": -184.994384765625,
+      "eval_loss": 0.36364680528640747,
+      "eval_rewards/accuracies": 0.8815789222717285,
+      "eval_rewards/chosen": 1.8633519411087036,
+      "eval_rewards/margins": 2.015190839767456,
+      "eval_rewards/rejected": -0.15183834731578827,
+      "eval_runtime": 27.7804,
+      "eval_samples_per_second": 10.835,
+      "eval_steps_per_second": 0.684,
+      "step": 100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 144.1908317487022,
+      "learning_rate": 9.991050648838675e-08,
+      "logits/chosen": -2.8495163917541504,
+      "logits/rejected": -2.819291353225708,
+      "logps/chosen": -62.78534698486328,
+      "logps/rejected": -203.07852172851562,
+      "loss": 0.3763,
+      "rewards/accuracies": 0.918749988079071,
+      "rewards/chosen": 1.7807117700576782,
+      "rewards/margins": 2.012101888656616,
+      "rewards/rejected": -0.231390118598938,
+      "step": 110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 126.7835721201166,
+      "learning_rate": 9.976379687325631e-08,
+      "logits/chosen": -2.7876718044281006,
+      "logits/rejected": -2.7896816730499268,
+      "logps/chosen": -63.9537353515625,
+      "logps/rejected": -183.58328247070312,
+      "loss": 0.3381,
+      "rewards/accuracies": 0.893750011920929,
+      "rewards/chosen": 1.7092759609222412,
+      "rewards/margins": 2.222316265106201,
+      "rewards/rejected": -0.5130404233932495,
+      "step": 120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 224.70875384146225,
+      "learning_rate": 9.954748808839674e-08,
+      "logits/chosen": -2.799767017364502,
+      "logits/rejected": -2.79805064201355,
+      "logps/chosen": -81.58202362060547,
+      "logps/rejected": -194.84872436523438,
+      "loss": 0.3302,
+      "rewards/accuracies": 0.84375,
+      "rewards/chosen": 1.278132438659668,
+      "rewards/margins": 2.2866179943084717,
+      "rewards/rejected": -1.0084853172302246,
+      "step": 130
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 152.12671314001338,
+      "learning_rate": 9.926188266120296e-08,
+      "logits/chosen": -2.763793706893921,
+      "logits/rejected": -2.727785348892212,
+      "logps/chosen": -75.18443298339844,
+      "logps/rejected": -176.11270141601562,
+      "loss": 0.3357,
+      "rewards/accuracies": 0.8500000238418579,
+      "rewards/chosen": 1.239394187927246,
+      "rewards/margins": 2.4168992042541504,
+      "rewards/rejected": -1.1775051355361938,
+      "step": 140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 175.5036541136684,
+      "learning_rate": 9.890738003669028e-08,
+      "logits/chosen": -2.7513179779052734,
+      "logits/rejected": -2.7361416816711426,
+      "logps/chosen": -81.78472900390625,
+      "logps/rejected": -219.29556274414062,
+      "loss": 0.2906,
+      "rewards/accuracies": 0.8812500238418579,
+      "rewards/chosen": 0.717334508895874,
+      "rewards/margins": 3.1458559036254883,
+      "rewards/rejected": -2.4285213947296143,
+      "step": 150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 170.33413139097263,
+      "learning_rate": 9.848447601883434e-08,
+      "logits/chosen": -2.6835436820983887,
+      "logits/rejected": -2.68257737159729,
+      "logps/chosen": -84.51789855957031,
+      "logps/rejected": -208.9514923095703,
+      "loss": 0.3133,
+      "rewards/accuracies": 0.893750011920929,
+      "rewards/chosen": 0.4816192090511322,
+      "rewards/margins": 2.897939443588257,
+      "rewards/rejected": -2.416320323944092,
+      "step": 160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 162.15836168812638,
+      "learning_rate": 9.799376207714445e-08,
+      "logits/chosen": -2.645550489425659,
+      "logits/rejected": -2.6841909885406494,
+      "logps/chosen": -82.82842254638672,
+      "logps/rejected": -195.99717712402344,
+      "loss": 0.3403,
+      "rewards/accuracies": 0.893750011920929,
+      "rewards/chosen": 0.4788424074649811,
+      "rewards/margins": 2.944272518157959,
+      "rewards/rejected": -2.46543025970459,
+      "step": 170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 119.29724112251517,
+      "learning_rate": 9.743592451943998e-08,
+      "logits/chosen": -2.687215805053711,
+      "logits/rejected": -2.6954283714294434,
+      "logps/chosen": -82.87139129638672,
+      "logps/rejected": -229.29202270507812,
+      "loss": 0.2932,
+      "rewards/accuracies": 0.8812500238418579,
+      "rewards/chosen": 0.5700880289077759,
+      "rewards/margins": 3.488856792449951,
+      "rewards/rejected": -2.9187686443328857,
+      "step": 180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 141.87377453577582,
+      "learning_rate": 9.681174353198685e-08,
+      "logits/chosen": -2.6703782081604004,
+      "logits/rejected": -2.6757917404174805,
+      "logps/chosen": -82.6243667602539,
+      "logps/rejected": -188.33477783203125,
+      "loss": 0.2942,
+      "rewards/accuracies": 0.8374999761581421,
+      "rewards/chosen": 0.5440769791603088,
+      "rewards/margins": 2.693603515625,
+      "rewards/rejected": -2.149526834487915,
+      "step": 190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 222.8132498459904,
+      "learning_rate": 9.612209208833647e-08,
+      "logits/chosen": -2.649033784866333,
+      "logits/rejected": -2.6577847003936768,
+      "logps/chosen": -78.76862335205078,
+      "logps/rejected": -221.69357299804688,
+      "loss": 0.2849,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.6831364631652832,
+      "rewards/margins": 3.6835713386535645,
+      "rewards/rejected": -3.000434398651123,
+      "step": 200
+    },
+    {
+      "epoch": 0.43,
+      "eval_logits/chosen": -2.5474884510040283,
+      "eval_logits/rejected": -2.5067243576049805,
+      "eval_logps/chosen": -75.19207763671875,
+      "eval_logps/rejected": -220.6973876953125,
+      "eval_loss": 0.2597656548023224,
+      "eval_rewards/accuracies": 0.8815789222717285,
+      "eval_rewards/chosen": 0.7705855369567871,
+      "eval_rewards/margins": 4.492726802825928,
+      "eval_rewards/rejected": -3.7221410274505615,
+      "eval_runtime": 27.7534,
+      "eval_samples_per_second": 10.846,
+      "eval_steps_per_second": 0.685,
+      "step": 200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 159.31516163900298,
+      "learning_rate": 9.536793472839323e-08,
+      "logits/chosen": -2.6742613315582275,
+      "logits/rejected": -2.6248650550842285,
+      "logps/chosen": -83.2725830078125,
+      "logps/rejected": -208.2786865234375,
+      "loss": 0.2658,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.7041572332382202,
+      "rewards/margins": 3.5826714038848877,
+      "rewards/rejected": -2.878513813018799,
+      "step": 210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 179.27000301013558,
+      "learning_rate": 9.455032620941839e-08,
+      "logits/chosen": -2.6498050689697266,
+      "logits/rejected": -2.676229953765869,
+      "logps/chosen": -89.39515686035156,
+      "logps/rejected": -219.5927276611328,
+      "loss": 0.3068,
+      "rewards/accuracies": 0.8687499761581421,
+      "rewards/chosen": 0.4539889395236969,
+      "rewards/margins": 3.6276097297668457,
+      "rewards/rejected": -3.1736209392547607,
+      "step": 220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 128.89336085629637,
+      "learning_rate": 9.367041003085648e-08,
+      "logits/chosen": -2.6379408836364746,
+      "logits/rejected": -2.672508716583252,
+      "logps/chosen": -81.60795593261719,
+      "logps/rejected": -219.58493041992188,
+      "loss": 0.247,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.5703738927841187,
+      "rewards/margins": 4.038177013397217,
+      "rewards/rejected": -3.4678032398223877,
+      "step": 230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 131.59397339189383,
+      "learning_rate": 9.272941683504808e-08,
+      "logits/chosen": -2.6790764331817627,
+      "logits/rejected": -2.6560444831848145,
+      "logps/chosen": -80.53921508789062,
+      "logps/rejected": -245.55868530273438,
+      "loss": 0.2733,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.6836718320846558,
+      "rewards/margins": 5.092911720275879,
+      "rewards/rejected": -4.409239768981934,
+      "step": 240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 172.8756884288556,
+      "learning_rate": 9.172866268606513e-08,
+      "logits/chosen": -2.6560401916503906,
+      "logits/rejected": -2.6558754444122314,
+      "logps/chosen": -83.98664855957031,
+      "logps/rejected": -211.24819946289062,
+      "loss": 0.2742,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 0.5306154489517212,
+      "rewards/margins": 3.6272902488708496,
+      "rewards/rejected": -3.096675157546997,
+      "step": 250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 133.84123014944902,
+      "learning_rate": 9.066954722907638e-08,
+      "logits/chosen": -2.6467976570129395,
+      "logits/rejected": -2.6493191719055176,
+      "logps/chosen": -81.35044860839844,
+      "logps/rejected": -215.6734161376953,
+      "loss": 0.2596,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 0.13885298371315002,
+      "rewards/margins": 3.929053783416748,
+      "rewards/rejected": -3.790201187133789,
+      "step": 260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 170.45984672890785,
+      "learning_rate": 8.955355173281707e-08,
+      "logits/chosen": -2.6520111560821533,
+      "logits/rejected": -2.662503719329834,
+      "logps/chosen": -92.13319396972656,
+      "logps/rejected": -233.948486328125,
+      "loss": 0.2782,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.021317768841981888,
+      "rewards/margins": 4.14536190032959,
+      "rewards/rejected": -4.124044418334961,
+      "step": 270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 229.80467359056394,
+      "learning_rate": 8.838223701790055e-08,
+      "logits/chosen": -2.6390814781188965,
+      "logits/rejected": -2.622087240219116,
+      "logps/chosen": -81.07563781738281,
+      "logps/rejected": -205.11767578125,
+      "loss": 0.2882,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 0.6619526147842407,
+      "rewards/margins": 3.7605528831481934,
+      "rewards/rejected": -3.098600387573242,
+      "step": 280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 141.1175297189335,
+      "learning_rate": 8.715724127386971e-08,
+      "logits/chosen": -2.631877899169922,
+      "logits/rejected": -2.6188888549804688,
+      "logps/chosen": -77.27508544921875,
+      "logps/rejected": -202.60848999023438,
+      "loss": 0.2766,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 0.7706214189529419,
+      "rewards/margins": 3.9430625438690186,
+      "rewards/rejected": -3.1724414825439453,
+      "step": 290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 128.57223622967007,
+      "learning_rate": 8.588027776804058e-08,
+      "logits/chosen": -2.659540891647339,
+      "logits/rejected": -2.6711039543151855,
+      "logps/chosen": -79.66828918457031,
+      "logps/rejected": -224.89675903320312,
+      "loss": 0.2496,
+      "rewards/accuracies": 0.9312499761581421,
+      "rewards/chosen": 0.8538252711296082,
+      "rewards/margins": 4.59223747253418,
+      "rewards/rejected": -3.7384121417999268,
+      "step": 300
+    },
+    {
+      "epoch": 0.64,
+      "eval_logits/chosen": -2.54941725730896,
+      "eval_logits/rejected": -2.507993698120117,
+      "eval_logps/chosen": -73.57530975341797,
+      "eval_logps/rejected": -226.1934356689453,
+      "eval_loss": 0.22951094806194305,
+      "eval_rewards/accuracies": 0.8684210777282715,
+      "eval_rewards/chosen": 0.9322623014450073,
+      "eval_rewards/margins": 5.204004287719727,
+      "eval_rewards/rejected": -4.271742343902588,
+      "eval_runtime": 27.7467,
+      "eval_samples_per_second": 10.848,
+      "eval_steps_per_second": 0.685,
+      "step": 300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 124.55913875585814,
+      "learning_rate": 8.455313244934323e-08,
+      "logits/chosen": -2.6776740550994873,
+      "logits/rejected": -2.6694130897521973,
+      "logps/chosen": -77.09771728515625,
+      "logps/rejected": -221.2405548095703,
+      "loss": 0.2385,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 0.8464420437812805,
+      "rewards/margins": 4.465517520904541,
+      "rewards/rejected": -3.619075059890747,
+      "step": 310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 126.82250398979915,
+      "learning_rate": 8.317766145051056e-08,
+      "logits/chosen": -2.655475616455078,
+      "logits/rejected": -2.646566867828369,
+      "logps/chosen": -90.94697570800781,
+      "logps/rejected": -239.02603149414062,
+      "loss": 0.2527,
+      "rewards/accuracies": 0.856249988079071,
+      "rewards/chosen": 0.3390585482120514,
+      "rewards/margins": 4.169254302978516,
+      "rewards/rejected": -3.830195903778076,
+      "step": 320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 128.9818135312364,
+      "learning_rate": 8.175578849210893e-08,
+      "logits/chosen": -2.628675699234009,
+      "logits/rejected": -2.63059401512146,
+      "logps/chosen": -79.80452728271484,
+      "logps/rejected": -214.23886108398438,
+      "loss": 0.2346,
+      "rewards/accuracies": 0.8812500238418579,
+      "rewards/chosen": 0.7122211456298828,
+      "rewards/margins": 4.485437393188477,
+      "rewards/rejected": -3.7732162475585938,
+      "step": 330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 170.62278854131603,
+      "learning_rate": 8.0289502192041e-08,
+      "logits/chosen": -2.6830954551696777,
+      "logits/rejected": -2.639918804168701,
+      "logps/chosen": -84.37899780273438,
+      "logps/rejected": -245.4980926513672,
+      "loss": 0.2358,
+      "rewards/accuracies": 0.893750011920929,
+      "rewards/chosen": 0.7535432577133179,
+      "rewards/margins": 4.955996990203857,
+      "rewards/rejected": -4.20245361328125,
+      "step": 340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 208.2810878095979,
+      "learning_rate": 7.878085328428368e-08,
+      "logits/chosen": -2.6215579509735107,
+      "logits/rejected": -2.6440749168395996,
+      "logps/chosen": -88.97381591796875,
+      "logps/rejected": -217.13919067382812,
+      "loss": 0.239,
+      "rewards/accuracies": 0.84375,
+      "rewards/chosen": 0.4338725507259369,
+      "rewards/margins": 3.943838596343994,
+      "rewards/rejected": -3.5099663734436035,
+      "step": 350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 129.5194204672045,
+      "learning_rate": 7.723195175075135e-08,
+      "logits/chosen": -2.633841037750244,
+      "logits/rejected": -2.6651291847229004,
+      "logps/chosen": -85.39970397949219,
+      "logps/rejected": -230.7396240234375,
+      "loss": 0.2339,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.3009163737297058,
+      "rewards/margins": 4.652897357940674,
+      "rewards/rejected": -4.351980686187744,
+      "step": 360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 164.1373245285756,
+      "learning_rate": 7.56449638702953e-08,
+      "logits/chosen": -2.614220142364502,
+      "logits/rejected": -2.6127521991729736,
+      "logps/chosen": -81.96704864501953,
+      "logps/rejected": -249.83847045898438,
+      "loss": 0.2353,
+      "rewards/accuracies": 0.893750011920929,
+      "rewards/chosen": 0.21380162239074707,
+      "rewards/margins": 4.993688106536865,
+      "rewards/rejected": -4.779886245727539,
+      "step": 370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 145.70422295065282,
+      "learning_rate": 7.402210918896688e-08,
+      "logits/chosen": -2.6128079891204834,
+      "logits/rejected": -2.6537294387817383,
+      "logps/chosen": -92.5042953491211,
+      "logps/rejected": -237.99649047851562,
+      "loss": 0.2447,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": -0.36580491065979004,
+      "rewards/margins": 4.624101638793945,
+      "rewards/rejected": -4.9899067878723145,
+      "step": 380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 187.73130049718608,
+      "learning_rate": 7.236565741578162e-08,
+      "logits/chosen": -2.6431612968444824,
+      "logits/rejected": -2.640349864959717,
+      "logps/chosen": -100.9793472290039,
+      "logps/rejected": -250.1385498046875,
+      "loss": 0.2144,
+      "rewards/accuracies": 0.856249988079071,
+      "rewards/chosen": -0.3357706665992737,
+      "rewards/margins": 4.967101097106934,
+      "rewards/rejected": -5.3028717041015625,
+      "step": 390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 149.7762972401691,
+      "learning_rate": 7.067792524832604e-08,
+      "logits/chosen": -2.6969475746154785,
+      "logits/rejected": -2.7058839797973633,
+      "logps/chosen": -87.51020812988281,
+      "logps/rejected": -230.3865966796875,
+      "loss": 0.2331,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 0.4354115426540375,
+      "rewards/margins": 4.516859531402588,
+      "rewards/rejected": -4.081448554992676,
+      "step": 400
+    },
+    {
+      "epoch": 0.86,
+      "eval_logits/chosen": -2.538165807723999,
+      "eval_logits/rejected": -2.4967117309570312,
+      "eval_logps/chosen": -75.02692413330078,
+      "eval_logps/rejected": -232.38836669921875,
+      "eval_loss": 0.2088860273361206,
+      "eval_rewards/accuracies": 0.8684210777282715,
+      "eval_rewards/chosen": 0.7871010899543762,
+      "eval_rewards/margins": 5.678337574005127,
+      "eval_rewards/rejected": -4.891236782073975,
+      "eval_runtime": 27.7296,
+      "eval_samples_per_second": 10.855,
+      "eval_steps_per_second": 0.685,
+      "step": 400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 156.8019073660933,
+      "learning_rate": 6.896127313264642e-08,
+      "logits/chosen": -2.64673113822937,
+      "logits/rejected": -2.7038655281066895,
+      "logps/chosen": -80.8321533203125,
+      "logps/rejected": -215.33200073242188,
+      "loss": 0.2199,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.6177021861076355,
+      "rewards/margins": 4.641745090484619,
+      "rewards/rejected": -4.02404260635376,
+      "step": 410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 159.07138805831667,
+      "learning_rate": 6.721810196195174e-08,
+      "logits/chosen": -2.6594576835632324,
+      "logits/rejected": -2.6509957313537598,
+      "logps/chosen": -85.29691314697266,
+      "logps/rejected": -228.0242156982422,
+      "loss": 0.2228,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 0.5911294221878052,
+      "rewards/margins": 4.615528583526611,
+      "rewards/rejected": -4.024399280548096,
+      "step": 420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 168.14810806211554,
+      "learning_rate": 6.545084971874738e-08,
+      "logits/chosen": -2.6609866619110107,
+      "logits/rejected": -2.6674180030822754,
+      "logps/chosen": -82.54444885253906,
+      "logps/rejected": -218.54458618164062,
+      "loss": 0.242,
+      "rewards/accuracies": 0.90625,
+      "rewards/chosen": 0.23233623802661896,
+      "rewards/margins": 4.690705299377441,
+      "rewards/rejected": -4.458369255065918,
+      "step": 430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 168.9123818369519,
+      "learning_rate": 6.3661988065096e-08,
+      "logits/chosen": -2.653125286102295,
+      "logits/rejected": -2.6496918201446533,
+      "logps/chosen": -90.27217102050781,
+      "logps/rejected": -240.4198455810547,
+      "loss": 0.2216,
+      "rewards/accuracies": 0.918749988079071,
+      "rewards/chosen": 0.1593785583972931,
+      "rewards/margins": 4.545279502868652,
+      "rewards/rejected": -4.38590145111084,
+      "step": 440
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 148.89537306570088,
+      "learning_rate": 6.185401888577488e-08,
+      "logits/chosen": -2.658726930618286,
+      "logits/rejected": -2.6689276695251465,
+      "logps/chosen": -80.66201782226562,
+      "logps/rejected": -221.6832733154297,
+      "loss": 0.2592,
+      "rewards/accuracies": 0.8500000238418579,
+      "rewards/chosen": 0.2890837788581848,
+      "rewards/margins": 4.448688507080078,
+      "rewards/rejected": -4.159604549407959,
+      "step": 450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 149.76478012255194,
+      "learning_rate": 6.002947078916364e-08,
+      "logits/chosen": -2.590343713760376,
+      "logits/rejected": -2.6157240867614746,
+      "logps/chosen": -88.16920471191406,
+      "logps/rejected": -197.1310577392578,
+      "loss": 0.2411,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": -0.010756964795291424,
+      "rewards/margins": 3.987351179122925,
+      "rewards/rejected": -3.998108386993408,
+      "step": 460
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 69.73653024843325,
+      "learning_rate": 5.8190895570756884e-08,
+      "logits/chosen": -2.6571004390716553,
+      "logits/rejected": -2.6537766456604004,
+      "logps/chosen": -91.45221710205078,
+      "logps/rejected": -221.7407684326172,
+      "loss": 0.1735,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.4764205813407898,
+      "rewards/margins": 5.178285121917725,
+      "rewards/rejected": -4.701864242553711,
+      "step": 470
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 74.69175649941268,
+      "learning_rate": 5.634086464424742e-08,
+      "logits/chosen": -2.629711866378784,
+      "logits/rejected": -2.628819227218628,
+      "logps/chosen": -81.51197052001953,
+      "logps/rejected": -247.3602294921875,
+      "loss": 0.1179,
+      "rewards/accuracies": 0.9937499761581421,
+      "rewards/chosen": 0.6573706865310669,
+      "rewards/margins": 5.840112209320068,
+      "rewards/rejected": -5.182741641998291,
+      "step": 480
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 81.17366180243518,
+      "learning_rate": 5.4481965445171675e-08,
+      "logits/chosen": -2.6255042552948,
+      "logits/rejected": -2.6369516849517822,
+      "logps/chosen": -92.74832153320312,
+      "logps/rejected": -253.36770629882812,
+      "loss": 0.1127,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.6330462098121643,
+      "rewards/margins": 5.925755500793457,
+      "rewards/rejected": -5.2927093505859375,
+      "step": 490
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 42.75765323253819,
+      "learning_rate": 5.261679781214719e-08,
+      "logits/chosen": -2.615849018096924,
+      "logits/rejected": -2.583253860473633,
+      "logps/chosen": -81.4043197631836,
+      "logps/rejected": -224.62063598632812,
+      "loss": 0.0874,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.7663254737854004,
+      "rewards/margins": 5.37575101852417,
+      "rewards/rejected": -4.6094255447387695,
+      "step": 500
+    },
+    {
+      "epoch": 1.07,
+      "eval_logits/chosen": -2.476076364517212,
+      "eval_logits/rejected": -2.4322872161865234,
+      "eval_logps/chosen": -76.55270385742188,
+      "eval_logps/rejected": -240.9202117919922,
+      "eval_loss": 0.1871984452009201,
+      "eval_rewards/accuracies": 0.8815789222717285,
+      "eval_rewards/chosen": 0.6345227956771851,
+      "eval_rewards/margins": 6.378945350646973,
+      "eval_rewards/rejected": -5.744422435760498,
+      "eval_runtime": 27.7245,
+      "eval_samples_per_second": 10.857,
+      "eval_steps_per_second": 0.685,
+      "step": 500
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 80.40588664774708,
+      "learning_rate": 5.074797035076318e-08,
+      "logits/chosen": -2.600032091140747,
+      "logits/rejected": -2.6238975524902344,
+      "logps/chosen": -80.40613555908203,
+      "logps/rejected": -222.7124786376953,
+      "loss": 0.1237,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.6139314770698547,
+      "rewards/margins": 5.547031402587891,
+      "rewards/rejected": -4.93310022354126,
+      "step": 510
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 144.58586359448068,
+      "learning_rate": 4.887809678520975e-08,
+      "logits/chosen": -2.5908000469207764,
+      "logits/rejected": -2.598961114883423,
+      "logps/chosen": -84.61862182617188,
+      "logps/rejected": -278.3133239746094,
+      "loss": 0.1324,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.5282140970230103,
+      "rewards/margins": 6.792176723480225,
+      "rewards/rejected": -6.263962745666504,
+      "step": 520
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 93.29727808293761,
+      "learning_rate": 4.7009792302748285e-08,
+      "logits/chosen": -2.5732998847961426,
+      "logits/rejected": -2.5646510124206543,
+      "logps/chosen": -93.49433898925781,
+      "logps/rejected": -266.72454833984375,
+      "loss": 0.1078,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.5050237774848938,
+      "rewards/margins": 6.757340908050537,
+      "rewards/rejected": -6.252317428588867,
+      "step": 530
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 100.11556094998792,
+      "learning_rate": 4.514566989613559e-08,
+      "logits/chosen": -2.5525121688842773,
+      "logits/rejected": -2.5795490741729736,
+      "logps/chosen": -82.06108856201172,
+      "logps/rejected": -232.47549438476562,
+      "loss": 0.0966,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.49808913469314575,
+      "rewards/margins": 5.77323579788208,
+      "rewards/rejected": -5.275146484375,
+      "step": 540
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 42.560167765291645,
+      "learning_rate": 4.328833670911724e-08,
+      "logits/chosen": -2.6239736080169678,
+      "logits/rejected": -2.5731518268585205,
+      "logps/chosen": -87.47773742675781,
+      "logps/rejected": -260.68145751953125,
+      "loss": 0.1202,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.3656201958656311,
+      "rewards/margins": 6.3458123207092285,
+      "rewards/rejected": -5.980191707611084,
+      "step": 550
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 93.5777273080985,
+      "learning_rate": 4.144039039010124e-08,
+      "logits/chosen": -2.5428924560546875,
+      "logits/rejected": -2.573913097381592,
+      "logps/chosen": -83.42646789550781,
+      "logps/rejected": -236.0640106201172,
+      "loss": 0.1036,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 0.4360501170158386,
+      "rewards/margins": 5.618575096130371,
+      "rewards/rejected": -5.182524681091309,
+      "step": 560
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 131.2737748435812,
+      "learning_rate": 3.960441545911204e-08,
+      "logits/chosen": -2.494786024093628,
+      "logits/rejected": -2.546558380126953,
+      "logps/chosen": -83.13180541992188,
+      "logps/rejected": -227.36331176757812,
+      "loss": 0.0986,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.48235464096069336,
+      "rewards/margins": 5.769989013671875,
+      "rewards/rejected": -5.28763484954834,
+      "step": 570
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 111.6736032390892,
+      "learning_rate": 3.778297969310529e-08,
+      "logits/chosen": -2.5785977840423584,
+      "logits/rejected": -2.6077938079833984,
+      "logps/chosen": -99.91156768798828,
+      "logps/rejected": -274.0140075683594,
+      "loss": 0.1194,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.2391074150800705,
+      "rewards/margins": 6.7299485206604,
+      "rewards/rejected": -6.490840911865234,
+      "step": 580
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 85.48495941913924,
+      "learning_rate": 3.5978630534699866e-08,
+      "logits/chosen": -2.5449576377868652,
+      "logits/rejected": -2.5609145164489746,
+      "logps/chosen": -79.68949127197266,
+      "logps/rejected": -262.533203125,
+      "loss": 0.1049,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.3723651170730591,
+      "rewards/margins": 6.530053615570068,
+      "rewards/rejected": -6.157688617706299,
+      "step": 590
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 113.71617513097341,
+      "learning_rate": 3.41938915293488e-08,
+      "logits/chosen": -2.5033059120178223,
+      "logits/rejected": -2.535037040710449,
+      "logps/chosen": -72.401123046875,
+      "logps/rejected": -242.56362915039062,
+      "loss": 0.1217,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.5488067865371704,
+      "rewards/margins": 6.211435794830322,
+      "rewards/rejected": -5.662629127502441,
+      "step": 600
+    },
+    {
+      "epoch": 1.28,
+      "eval_logits/chosen": -2.417189359664917,
+      "eval_logits/rejected": -2.374096632003784,
+      "eval_logps/chosen": -80.61611938476562,
+      "eval_logps/rejected": -250.38265991210938,
+      "eval_loss": 0.1832417994737625,
+      "eval_rewards/accuracies": 0.8684210777282715,
+      "eval_rewards/chosen": 0.22818052768707275,
+      "eval_rewards/margins": 6.918849945068359,
+      "eval_rewards/rejected": -6.690669536590576,
+      "eval_runtime": 27.7688,
+      "eval_samples_per_second": 10.84,
+      "eval_steps_per_second": 0.684,
+      "step": 600
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 72.12098063143571,
+      "learning_rate": 3.2431258795932856e-08,
+      "logits/chosen": -2.5755138397216797,
+      "logits/rejected": -2.567202568054199,
+      "logps/chosen": -85.8843002319336,
+      "logps/rejected": -263.94622802734375,
+      "loss": 0.0791,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.24374854564666748,
+      "rewards/margins": 6.808441162109375,
+      "rewards/rejected": -6.564692497253418,
+      "step": 610
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 84.30497343810121,
+      "learning_rate": 3.0693197535712687e-08,
+      "logits/chosen": -2.556939125061035,
+      "logits/rejected": -2.553295373916626,
+      "logps/chosen": -93.59156036376953,
+      "logps/rejected": -255.7102813720703,
+      "loss": 0.104,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.050883639603853226,
+      "rewards/margins": 6.550814628601074,
+      "rewards/rejected": -6.601698398590088,
+      "step": 620
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 97.92510560815347,
+      "learning_rate": 2.8982138584521732e-08,
+      "logits/chosen": -2.523881435394287,
+      "logits/rejected": -2.5281319618225098,
+      "logps/chosen": -87.27870178222656,
+      "logps/rejected": -281.7742919921875,
+      "loss": 0.0988,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.02099490724503994,
+      "rewards/margins": 6.995011329650879,
+      "rewards/rejected": -6.974015712738037,
+      "step": 630
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 144.6029547590291,
+      "learning_rate": 2.730047501302266e-08,
+      "logits/chosen": -2.511509656906128,
+      "logits/rejected": -2.491812229156494,
+      "logps/chosen": -80.01902770996094,
+      "logps/rejected": -249.27377319335938,
+      "loss": 0.1017,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.18679390847682953,
+      "rewards/margins": 6.189632415771484,
+      "rewards/rejected": -6.002839088439941,
+      "step": 640
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 51.26390219844907,
+      "learning_rate": 2.5650558779781632e-08,
+      "logits/chosen": -2.468006134033203,
+      "logits/rejected": -2.5272440910339355,
+      "logps/chosen": -92.99039459228516,
+      "logps/rejected": -263.6413269042969,
+      "loss": 0.1057,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.007961297407746315,
+      "rewards/margins": 6.804115295410156,
+      "rewards/rejected": -6.7961530685424805,
+      "step": 650
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 74.26596525719256,
+      "learning_rate": 2.403469744184154e-08,
+      "logits/chosen": -2.4938857555389404,
+      "logits/rejected": -2.4678053855895996,
+      "logps/chosen": -83.55183410644531,
+      "logps/rejected": -240.530517578125,
+      "loss": 0.0828,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.22306349873542786,
+      "rewards/margins": 7.2999396324157715,
+      "rewards/rejected": -7.076876640319824,
+      "step": 660
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 67.9457822191306,
+      "learning_rate": 2.2455150927394877e-08,
+      "logits/chosen": -2.5248875617980957,
+      "logits/rejected": -2.524660587310791,
+      "logps/chosen": -91.45790100097656,
+      "logps/rejected": -274.40240478515625,
+      "loss": 0.1147,
+      "rewards/accuracies": 0.956250011920929,
+      "rewards/chosen": -0.18505340814590454,
+      "rewards/margins": 6.406013488769531,
+      "rewards/rejected": -6.591066837310791,
+      "step": 670
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 93.08918649012821,
+      "learning_rate": 2.0914128375069722e-08,
+      "logits/chosen": -2.503322124481201,
+      "logits/rejected": -2.5299384593963623,
+      "logps/chosen": -87.551025390625,
+      "logps/rejected": -258.9871520996094,
+      "loss": 0.0997,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.09187181293964386,
+      "rewards/margins": 6.872333526611328,
+      "rewards/rejected": -6.78046178817749,
+      "step": 680
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 47.52507218664577,
+      "learning_rate": 1.9413785044249675e-08,
+      "logits/chosen": -2.4908030033111572,
+      "logits/rejected": -2.4941422939300537,
+      "logps/chosen": -92.09065246582031,
+      "logps/rejected": -250.3391876220703,
+      "loss": 0.0975,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.06340374052524567,
+      "rewards/margins": 6.6621413230896,
+      "rewards/rejected": -6.598738193511963,
+      "step": 690
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 165.96896499860316,
+      "learning_rate": 1.795621930074879e-08,
+      "logits/chosen": -2.4552981853485107,
+      "logits/rejected": -2.4628846645355225,
+      "logps/chosen": -87.13786315917969,
+      "logps/rejected": -213.85501098632812,
+      "loss": 0.0966,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.18398265540599823,
+      "rewards/margins": 5.881737232208252,
+      "rewards/rejected": -5.697754383087158,
+      "step": 700
+    },
+    {
+      "epoch": 1.5,
+      "eval_logits/chosen": -2.393958330154419,
+      "eval_logits/rejected": -2.3503360748291016,
+      "eval_logps/chosen": -81.04850006103516,
+      "eval_logps/rejected": -253.6011962890625,
+      "eval_loss": 0.18069878220558167,
+      "eval_rewards/accuracies": 0.8815789222717285,
+      "eval_rewards/chosen": 0.18494324386119843,
+      "eval_rewards/margins": 7.197464942932129,
+      "eval_rewards/rejected": -7.012521743774414,
+      "eval_runtime": 27.7257,
+      "eval_samples_per_second": 10.856,
+      "eval_steps_per_second": 0.685,
+      "step": 700
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 159.91079268011805,
+      "learning_rate": 1.6543469682057103e-08,
+      "logits/chosen": -2.5185043811798096,
+      "logits/rejected": -2.4764022827148438,
+      "logps/chosen": -84.46337890625,
+      "logps/rejected": -264.3829345703125,
+      "loss": 0.0926,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.18679778277873993,
+      "rewards/margins": 7.148007392883301,
+      "rewards/rejected": -6.961209297180176,
+      "step": 710
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 48.047361074216745,
+      "learning_rate": 1.5177512046261666e-08,
+      "logits/chosen": -2.484215259552002,
+      "logits/rejected": -2.507033109664917,
+      "logps/chosen": -84.62954711914062,
+      "logps/rejected": -256.36737060546875,
+      "loss": 0.0983,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.39748048782348633,
+      "rewards/margins": 7.362854957580566,
+      "rewards/rejected": -6.965374946594238,
+      "step": 720
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 112.9192400527363,
+      "learning_rate": 1.3860256808630427e-08,
+      "logits/chosen": -2.555844783782959,
+      "logits/rejected": -2.478036403656006,
+      "logps/chosen": -88.8625259399414,
+      "logps/rejected": -237.5943145751953,
+      "loss": 0.107,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.22078856825828552,
+      "rewards/margins": 6.411069393157959,
+      "rewards/rejected": -6.190280914306641,
+      "step": 730
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 80.75197492702765,
+      "learning_rate": 1.2593546269723649e-08,
+      "logits/chosen": -2.4870190620422363,
+      "logits/rejected": -2.470374584197998,
+      "logps/chosen": -87.91392517089844,
+      "logps/rejected": -268.05450439453125,
+      "loss": 0.103,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.023371051996946335,
+      "rewards/margins": 7.2978057861328125,
+      "rewards/rejected": -7.274435520172119,
+      "step": 740
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 108.88349531511021,
+      "learning_rate": 1.1379152038770029e-08,
+      "logits/chosen": -2.498399019241333,
+      "logits/rejected": -2.522373676300049,
+      "logps/chosen": -79.47743225097656,
+      "logps/rejected": -267.198486328125,
+      "loss": 0.1003,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.3667542338371277,
+      "rewards/margins": 7.25054931640625,
+      "rewards/rejected": -6.883795261383057,
+      "step": 750
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 106.04443890243027,
+      "learning_rate": 1.0218772555910953e-08,
+      "logits/chosen": -2.49924898147583,
+      "logits/rejected": -2.533095121383667,
+      "logps/chosen": -88.90888977050781,
+      "logps/rejected": -216.93911743164062,
+      "loss": 0.1093,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.10971729457378387,
+      "rewards/margins": 5.331303596496582,
+      "rewards/rejected": -5.221585750579834,
+      "step": 760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 119.24007089314645,
+      "learning_rate": 9.114030716778432e-09,
+      "logits/chosen": -2.522010087966919,
+      "logits/rejected": -2.510510206222534,
+      "logps/chosen": -85.21174621582031,
+      "logps/rejected": -258.8554992675781,
+      "loss": 0.0854,
+      "rewards/accuracies": 0.981249988079071,
+      "rewards/chosen": 0.3466641306877136,
+      "rewards/margins": 7.279857635498047,
+      "rewards/rejected": -6.933194160461426,
+      "step": 770
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 134.5502526352079,
+      "learning_rate": 8.066471602728803e-09,
+      "logits/chosen": -2.4917855262756348,
+      "logits/rejected": -2.505199670791626,
+      "logps/chosen": -90.26045989990234,
+      "logps/rejected": -252.13491821289062,
+      "loss": 0.085,
+      "rewards/accuracies": 0.956250011920929,
+      "rewards/chosen": 0.11037313938140869,
+      "rewards/margins": 6.754474639892578,
+      "rewards/rejected": -6.644102573394775,
+      "step": 780
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 70.79957881261483,
+      "learning_rate": 7.077560319906695e-09,
+      "logits/chosen": -2.4431352615356445,
+      "logits/rejected": -2.495988368988037,
+      "logps/chosen": -79.27693176269531,
+      "logps/rejected": -225.5204620361328,
+      "loss": 0.0872,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.49195781350135803,
+      "rewards/margins": 6.213925361633301,
+      "rewards/rejected": -5.721967697143555,
+      "step": 790
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 111.87130217318148,
+      "learning_rate": 6.148679950161672e-09,
+      "logits/chosen": -2.475398302078247,
+      "logits/rejected": -2.5179803371429443,
+      "logps/chosen": -89.04031372070312,
+      "logps/rejected": -270.52117919921875,
+      "loss": 0.0755,
+      "rewards/accuracies": 0.981249988079071,
+      "rewards/chosen": 0.1288180649280548,
+      "rewards/margins": 7.451496124267578,
+      "rewards/rejected": -7.322678565979004,
+      "step": 800
+    },
+    {
+      "epoch": 1.71,
+      "eval_logits/chosen": -2.3867335319519043,
+      "eval_logits/rejected": -2.3436965942382812,
+      "eval_logps/chosen": -79.67388153076172,
+      "eval_logps/rejected": -253.01498413085938,
+      "eval_loss": 0.18022477626800537,
+      "eval_rewards/accuracies": 0.8947368264198303,
+      "eval_rewards/chosen": 0.32240524888038635,
+      "eval_rewards/margins": 7.276304721832275,
+      "eval_rewards/rejected": -6.953899383544922,
+      "eval_runtime": 27.7976,
+      "eval_samples_per_second": 10.828,
+      "eval_steps_per_second": 0.684,
+      "step": 800
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 110.49991026961725,
+      "learning_rate": 5.281129616683166e-09,
+      "logits/chosen": -2.5127198696136475,
+      "logits/rejected": -2.512899160385132,
+      "logps/chosen": -86.99423217773438,
+      "logps/rejected": -254.9462432861328,
+      "loss": 0.0962,
+      "rewards/accuracies": 0.9437500238418579,
+      "rewards/chosen": 0.06987112015485764,
+      "rewards/margins": 6.7237701416015625,
+      "rewards/rejected": -6.6538987159729,
+      "step": 810
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 108.21879194257498,
+      "learning_rate": 4.476122667059206e-09,
+      "logits/chosen": -2.4863157272338867,
+      "logits/rejected": -2.4990665912628174,
+      "logps/chosen": -86.436279296875,
+      "logps/rejected": -256.197998046875,
+      "loss": 0.1138,
+      "rewards/accuracies": 0.981249988079071,
+      "rewards/chosen": 0.21532371640205383,
+      "rewards/margins": 6.994332313537598,
+      "rewards/rejected": -6.779008388519287,
+      "step": 820
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 76.19186603851071,
+      "learning_rate": 3.734784976300165e-09,
+      "logits/chosen": -2.4668314456939697,
+      "logits/rejected": -2.4680819511413574,
+      "logps/chosen": -89.12982940673828,
+      "logps/rejected": -259.98974609375,
+      "loss": 0.1034,
+      "rewards/accuracies": 0.956250011920929,
+      "rewards/chosen": 0.2635418772697449,
+      "rewards/margins": 6.681604862213135,
+      "rewards/rejected": -6.418063163757324,
+      "step": 830
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 89.59280146601071,
+      "learning_rate": 3.0581533722006947e-09,
+      "logits/chosen": -2.521101236343384,
+      "logits/rejected": -2.5018861293792725,
+      "logps/chosen": -85.35905456542969,
+      "logps/rejected": -227.835205078125,
+      "loss": 0.082,
+      "rewards/accuracies": 0.9937499761581421,
+      "rewards/chosen": 0.3482588827610016,
+      "rewards/margins": 6.087993144989014,
+      "rewards/rejected": -5.739733695983887,
+      "step": 840
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 132.74610726242307,
+      "learning_rate": 2.4471741852423233e-09,
+      "logits/chosen": -2.4352290630340576,
+      "logits/rejected": -2.490058422088623,
+      "logps/chosen": -89.67268371582031,
+      "logps/rejected": -224.5243682861328,
+      "loss": 0.1031,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.25209230184555054,
+      "rewards/margins": 6.445871829986572,
+      "rewards/rejected": -6.1937785148620605,
+      "step": 850
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 95.32823058547511,
+      "learning_rate": 1.9027019250647036e-09,
+      "logits/chosen": -2.5332539081573486,
+      "logits/rejected": -2.539168357849121,
+      "logps/chosen": -89.87360382080078,
+      "logps/rejected": -267.892822265625,
+      "loss": 0.0952,
+      "rewards/accuracies": 0.981249988079071,
+      "rewards/chosen": 0.2068900316953659,
+      "rewards/margins": 6.962863922119141,
+      "rewards/rejected": -6.755973815917969,
+      "step": 860
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 145.27755044968825,
+      "learning_rate": 1.4254980853566246e-09,
+      "logits/chosen": -2.4520225524902344,
+      "logits/rejected": -2.4698426723480225,
+      "logps/chosen": -85.50424194335938,
+      "logps/rejected": -255.6893768310547,
+      "loss": 0.0956,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.2579875588417053,
+      "rewards/margins": 7.30988073348999,
+      "rewards/rejected": -7.0518927574157715,
+      "step": 870
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 87.4062367020671,
+      "learning_rate": 1.0162300788382261e-09,
+      "logits/chosen": -2.488208770751953,
+      "logits/rejected": -2.515857458114624,
+      "logps/chosen": -90.32164001464844,
+      "logps/rejected": -226.85171508789062,
+      "loss": 0.0942,
+      "rewards/accuracies": 0.96875,
+      "rewards/chosen": 0.23069357872009277,
+      "rewards/margins": 6.096683502197266,
+      "rewards/rejected": -5.865989685058594,
+      "step": 880
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 129.60497880498312,
+      "learning_rate": 6.754703038239329e-10,
+      "logits/chosen": -2.4546942710876465,
+      "logits/rejected": -2.498014450073242,
+      "logps/chosen": -93.2281494140625,
+      "logps/rejected": -241.6513671875,
+      "loss": 0.0995,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.2960784137248993,
+      "rewards/margins": 6.222127914428711,
+      "rewards/rejected": -5.926049709320068,
+      "step": 890
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 189.4146678200498,
+      "learning_rate": 4.0369534367168944e-10,
+      "logits/chosen": -2.4957008361816406,
+      "logits/rejected": -2.4675281047821045,
+      "logps/chosen": -77.89246368408203,
+      "logps/rejected": -257.5621643066406,
+      "loss": 0.1177,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.49517494440078735,
+      "rewards/margins": 7.060118198394775,
+      "rewards/rejected": -6.564943790435791,
+      "step": 900
+    },
+    {
+      "epoch": 1.93,
+      "eval_logits/chosen": -2.384641647338867,
+      "eval_logits/rejected": -2.3416695594787598,
+      "eval_logps/chosen": -80.31053924560547,
+      "eval_logps/rejected": -253.77731323242188,
+      "eval_loss": 0.17933423817157745,
+      "eval_rewards/accuracies": 0.8947368264198303,
+      "eval_rewards/chosen": 0.25873973965644836,
+      "eval_rewards/margins": 7.288874626159668,
+      "eval_rewards/rejected": -7.030135154724121,
+      "eval_runtime": 27.722,
+      "eval_samples_per_second": 10.858,
+      "eval_steps_per_second": 0.685,
+      "step": 900
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 122.05899501763088,
+      "learning_rate": 2.0128530023804656e-10,
+      "logits/chosen": -2.496187448501587,
+      "logits/rejected": -2.4914169311523438,
+      "logps/chosen": -88.0294189453125,
+      "logps/rejected": -257.91558837890625,
+      "loss": 0.1067,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.12927184998989105,
+      "rewards/margins": 6.56699275970459,
+      "rewards/rejected": -6.437721252441406,
+      "step": 910
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 65.9343080763714,
+      "learning_rate": 6.852326227130833e-11,
+      "logits/chosen": -2.521149158477783,
+      "logits/rejected": -2.5130820274353027,
+      "logps/chosen": -84.37642669677734,
+      "logps/rejected": -246.3243865966797,
+      "loss": 0.0899,
+      "rewards/accuracies": 0.9937499761581421,
+      "rewards/chosen": 0.4820326268672943,
+      "rewards/margins": 6.372977256774902,
+      "rewards/rejected": -5.890944957733154,
+      "step": 920
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 93.45308900990692,
+      "learning_rate": 5.594909486328347e-12,
+      "logits/chosen": -2.450763702392578,
+      "logits/rejected": -2.467440128326416,
+      "logps/chosen": -78.51066589355469,
+      "logps/rejected": -229.6824188232422,
+      "loss": 0.0926,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.3496243357658386,
+      "rewards/margins": 5.816868305206299,
+      "rewards/rejected": -5.4672441482543945,
+      "step": 930
+    },
+    {
+      "epoch": 2.0,
+      "step": 934,
+      "total_flos": 0.0,
+      "train_loss": 0.21607095603003493,
+      "train_runtime": 14362.1771,
+      "train_samples_per_second": 4.163,
+      "train_steps_per_second": 0.065
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 934,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 50,
+  "total_flos": 0.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:671b65610da39dbaa29fbb27313c65bc3ff72cb5c681dbc518d28d001dda8d40
+size 6456