End of training

README.md

@@ -3,9 +3,17 @@ library_name: transformers
 license: apache-2.0
 base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- data/ui_math
+- data/ui_coding
+- data/ui_logic
 model-index:
 - name: tinyllama-1.1b-chat-v1.0-ui-dpo-2
   results: []
@@ -16,7 +24,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tinyllama-1.1b-chat-v1.0-ui-dpo-2
 
-This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) on the None dataset.
+This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) on the data/ui_math, the data/ui_coding and the data/ui_logic datasets.
+It achieves the following results on the evaluation set:
+- Loss: 0.6931
+- Rewards/chosen: 0.0
+- Rewards/rejected: 0.0
+- Rewards/accuracies: 0.0
+- Rewards/margins: 0.0
+- Logps/rejected: -239.1279
+- Logps/chosen: -225.0590
+- Logits/rejected: -2.3130
+- Logits/chosen: -2.1421
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.0,
+    "eval_logits/chosen": -2.142118453979492,
+    "eval_logits/rejected": -2.312952756881714,
+    "eval_logps/chosen": -225.0589599609375,
+    "eval_logps/rejected": -239.12791442871094,
+    "eval_loss": 0.6931471228599548,
+    "eval_rewards/accuracies": 0.0,
+    "eval_rewards/chosen": 0.0,
+    "eval_rewards/margins": 0.0,
+    "eval_rewards/rejected": 0.0,
+    "eval_runtime": 0.76,
+    "eval_samples": 50,
+    "eval_samples_per_second": 65.79,
+    "eval_steps_per_second": 1.316,
     "total_flos": 0.0,
     "train_loss": 0.1732867956161499,
     "train_runtime": 32.0764,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.1",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.0,
+    "eval_logits/chosen": -2.142118453979492,
+    "eval_logits/rejected": -2.312952756881714,
+    "eval_logps/chosen": -225.0589599609375,
+    "eval_logps/rejected": -239.12791442871094,
+    "eval_loss": 0.6931471228599548,
+    "eval_rewards/accuracies": 0.0,
+    "eval_rewards/chosen": 0.0,
+    "eval_rewards/margins": 0.0,
+    "eval_rewards/rejected": 0.0,
+    "eval_runtime": 0.76,
+    "eval_samples": 50,
+    "eval_samples_per_second": 65.79,
+    "eval_steps_per_second": 1.316
+}