End of training

README.md

@@ -2,10 +2,13 @@
 license: apache-2.0
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
 base_model: alignment-handbook/zephyr-7b-sft-full
+datasets:
+- generation/GPT4
 model-index:
 - name: zephyr-dpo-qlora-gpt4-5e-6-epoch3
   results: []
@@ -16,20 +19,20 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-dpo-qlora-gpt4-5e-6-epoch3
 
-This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the None dataset.
+This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the generation/GPT4 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.8351
-- Rewards/chosen: -16.6710
-- Rewards/rejected: -19.6029
-- Rewards/accuracies: 0.6825
-- Rewards/margins: 2.9319
-- Rewards/margins Max: 11.9629
-- Rewards/margins Min: -5.2899
-- Rewards/margins Std: 7.7662
-- Logps/rejected: -2219.4746
-- Logps/chosen: -1952.3245
-- Logits/rejected: -1.4296
-- Logits/chosen: -1.5156
+- Loss: 1.8389
+- Rewards/chosen: -16.6323
+- Rewards/rejected: -19.5704
+- Rewards/accuracies: 0.6905
+- Rewards/margins: 2.9381
+- Rewards/margins Max: 11.9980
+- Rewards/margins Min: -5.2949
+- Rewards/margins Std: 7.7821
+- Logps/rejected: -2216.2207
+- Logps/chosen: -1948.4508
+- Logits/rejected: -1.4331
+- Logits/chosen: -1.5188
 
 ## Model description
 

all_results.json

@@ -1,5 +1,21 @@
 {
     "epoch": 3.0,
+    "eval_logits/chosen": -1.5187865495681763,
+    "eval_logits/rejected": -1.4330620765686035,
+    "eval_logps/chosen": -1948.4508056640625,
+    "eval_logps/rejected": -2216.220703125,
+    "eval_loss": 1.8388876914978027,
+    "eval_rewards/accuracies": 0.6904761791229248,
+    "eval_rewards/chosen": -16.632295608520508,
+    "eval_rewards/margins": 2.9380877017974854,
+    "eval_rewards/margins_max": 11.99801254272461,
+    "eval_rewards/margins_min": -5.294867038726807,
+    "eval_rewards/margins_std": 7.78206729888916,
+    "eval_rewards/rejected": -19.570384979248047,
+    "eval_runtime": 281.8693,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 7.095,
+    "eval_steps_per_second": 0.224,
     "train_loss": 0.1103198329137612,
     "train_runtime": 9245.0119,
     "train_samples": 5678,

config.json

@@ -0,0 +1,40 @@
+{
+  "_name_or_path": "alignment-handbook/zephyr-7b-sft-full",
+  "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,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": false,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000.0,
+  "sliding_window": 4096,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.39.0.dev0",
+  "use_cache": true,
+  "vocab_size": 32000
+}

eval_results.json

@@ -0,0 +1,19 @@
+{
+    "epoch": 3.0,
+    "eval_logits/chosen": -1.5187865495681763,
+    "eval_logits/rejected": -1.4330620765686035,
+    "eval_logps/chosen": -1948.4508056640625,
+    "eval_logps/rejected": -2216.220703125,
+    "eval_loss": 1.8388876914978027,
+    "eval_rewards/accuracies": 0.6904761791229248,
+    "eval_rewards/chosen": -16.632295608520508,
+    "eval_rewards/margins": 2.9380877017974854,
+    "eval_rewards/margins_max": 11.99801254272461,
+    "eval_rewards/margins_min": -5.294867038726807,
+    "eval_rewards/margins_std": 7.78206729888916,
+    "eval_rewards/rejected": -19.570384979248047,
+    "eval_runtime": 281.8693,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 7.095,
+    "eval_steps_per_second": 0.224
+}

runs/Jul29_11-56-09_node12/events.out.tfevents.1722231834.node12.2246976.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9889ad12d5be2151c425c32d7ab4c23dda000613126770a218837f1bf71a915a
+size 1017