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-dpop-qlora-gpt4-5e-7-epoch3
   results: []
@@ -16,22 +19,22 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-dpop-qlora-gpt4-5e-7-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: 1.5707
-- Positive Losses: 8.9081
-- Dpo Losses: 0.6660
-- Rewards/chosen: -0.0431
-- Rewards/rejected: -0.1110
+- Loss: 1.5678
+- Positive Losses: 8.8531
+- Dpo Losses: 0.6658
+- Rewards/chosen: -0.0421
+- Rewards/rejected: -0.1106
 - Rewards/accuracies: 0.6151
-- Rewards/margins: 0.0679
-- Rewards/margins Max: 0.3167
-- Rewards/margins Min: -0.1568
-- Rewards/margins Std: 0.2111
-- Logps/rejected: -270.2825
-- Logps/chosen: -289.5273
-- Logits/rejected: -2.6606
-- Logits/chosen: -2.7037
+- Rewards/margins: 0.0684
+- Rewards/margins Max: 0.3178
+- Rewards/margins Min: -0.1570
+- Rewards/margins Std: 0.2114
+- Logps/rejected: -270.2393
+- Logps/chosen: -289.4331
+- Logits/rejected: -2.6601
+- Logits/chosen: -2.7033
 
 ## Model description
 

all_results.json

@@ -1,5 +1,23 @@
 {
     "epoch": 3.0,
+    "eval_dpo_losses": 0.6657702922821045,
+    "eval_logits/chosen": -2.7033236026763916,
+    "eval_logits/rejected": -2.660062789916992,
+    "eval_logps/chosen": -289.4331359863281,
+    "eval_logps/rejected": -270.23931884765625,
+    "eval_loss": 1.5678311586380005,
+    "eval_positive_losses": 8.853089332580566,
+    "eval_rewards/accuracies": 0.6150793433189392,
+    "eval_rewards/chosen": -0.04211931675672531,
+    "eval_rewards/margins": 0.0684470534324646,
+    "eval_rewards/margins_max": 0.31783527135849,
+    "eval_rewards/margins_min": -0.1569632887840271,
+    "eval_rewards/margins_std": 0.2114361971616745,
+    "eval_rewards/rejected": -0.11056637018918991,
+    "eval_runtime": 345.2654,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 5.793,
+    "eval_steps_per_second": 0.182,
     "train_loss": 0.5596254680078354,
     "train_runtime": 11165.5936,
     "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 +1,21 @@
+{
+    "epoch": 3.0,
+    "eval_dpo_losses": 0.6657702922821045,
+    "eval_logits/chosen": -2.7033236026763916,
+    "eval_logits/rejected": -2.660062789916992,
+    "eval_logps/chosen": -289.4331359863281,
+    "eval_logps/rejected": -270.23931884765625,
+    "eval_loss": 1.5678311586380005,
+    "eval_positive_losses": 8.853089332580566,
+    "eval_rewards/accuracies": 0.6150793433189392,
+    "eval_rewards/chosen": -0.04211931675672531,
+    "eval_rewards/margins": 0.0684470534324646,
+    "eval_rewards/margins_max": 0.31783527135849,
+    "eval_rewards/margins_min": -0.1569632887840271,
+    "eval_rewards/margins_std": 0.2114361971616745,
+    "eval_rewards/rejected": -0.11056637018918991,
+    "eval_runtime": 345.2654,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 5.793,
+    "eval_steps_per_second": 0.182
+}

runs/Jul29_12-00-30_node02/events.out.tfevents.1722234022.node02.3000485.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9238acc8f18ee333eea2911e33d3208e3f587d6a731785d1c88efb5781cd962
+size 1130