初始化项目，由ModelHub XC社区提供模型

Model: jackf857/qwen3-8b-base-beta-dpo-hh-harmless-4xh200-batch-64-20260424-025105 Source: Original Platform
2026-05-16 07:03:57 +08:00
commit 527c912d5f
23 changed files with 163379 additions and 0 deletions
--- a/README.md
+++ b/README.md
@@ -0,0 +1,79 @@
+---
+library_name: transformers
+base_model: jackf857/qwen3-8b-base-sft-hh-harmless-4xh200-batch-64-20260417-214452
+tags:
+- alignment-handbook
+- beta-dpo
+- generated_from_trainer
+datasets:
+- Anthropic/hh-rlhf
+model-index:
+- name: qwen3-8b-base-beta-dpo-hh-harmless-4xh200-batch-64-20260424-025105
+  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. -->
+
+# qwen3-8b-base-beta-dpo-hh-harmless-4xh200-batch-64-20260424-025105
+
+This model is a fine-tuned version of [jackf857/qwen3-8b-base-sft-hh-harmless-4xh200-batch-64-20260417-214452](https://huggingface.co/jackf857/qwen3-8b-base-sft-hh-harmless-4xh200-batch-64-20260417-214452) on the Anthropic/hh-rlhf dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.7256
+- Beta Dpo/gap Mean: 9.9202
+- Beta Dpo/gap Std: 18.3470
+- Beta Dpo/beta Used Raw: 0.1809
+- Beta Dpo/beta Used: 0.1995
+- Beta Dpo/mask Keep Frac: 1.0
+- Logits/chosen: 1.5449
+- Logits/rejected: 1.4137
+
+## 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: 5e-07
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 64
+- total_eval_batch_size: 32
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss | Beta Dpo/gap Mean | Beta Dpo/gap Std | Beta Dpo/beta Used Raw | Beta Dpo/beta Used | Beta Dpo/mask Keep Frac | Logits/chosen | Logits/rejected |
+|:-------------:|:------:|:----:|:---------------:|:-----------------:|:----------------:|:----------------------:|:------------------:|:-----------------------:|:-------------:|:---------------:|
+| 1.3382        | 0.1512 | 100  | 0.6596          | 0.5103            | 1.3374           | 0.1093                 | 0.1093             | 1.0                     | 1.8052        | 1.6946          |
+| 1.0452        | 0.3023 | 200  | 0.6042          | 5.1820            | 10.9635          | 0.1250                 | 0.1302             | 1.0                     | 1.6393        | 1.5121          |
+| 1.1502        | 0.4535 | 300  | 0.6283          | 8.5454            | 15.3857          | 0.1243                 | 0.1420             | 1.0                     | 1.4622        | 1.3384          |
+| 1.3806        | 0.6047 | 400  | 0.6464          | 9.9655            | 16.5703          | 0.1189                 | 0.1405             | 1.0                     | 1.4474        | 1.3215          |
+| 1.3396        | 0.7559 | 500  | 0.6756          | 11.0204           | 19.5206          | 0.1269                 | 0.1533             | 1.0                     | 1.3984        | 1.2735          |
+| 1.0636        | 0.9070 | 600  | 0.7256          | 9.9202            | 18.3470          | 0.1809                 | 0.1995             | 1.0                     | 1.5449        | 1.4137          |
+
+
+### Framework versions
+
+- Transformers 4.51.0
+- Pytorch 2.3.1+cu121
+- Datasets 2.21.0
+- Tokenizers 0.21.4