End of training

README.md

@@ -2,9 +2,15 @@
 license: apache-2.0
 base_model: martimfasantos/tinyllama-1.1b-sum-sft-full_old
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- openai/summarize_from_feedback
 model-index:
 - name: tinyllama-1.1b-sum-dpo-full_LR2e-8_3epochs_old
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tinyllama-1.1b-sum-dpo-full_LR2e-8_3epochs_old
 
-This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full_old](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full_old) on an unknown dataset.
+This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full_old](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full_old) on the openai/summarize_from_feedback dataset.
 It achieves the following results on the evaluation set:
-- Logits/chosen: -3.0704
-- Logits/rejected: -3.0648
-- Logps/chosen: -62.9434
-- Logps/rejected: -68.7246
 - Loss: 0.6872
-- Rewards/accuracies: 0.5943
 - Rewards/chosen: -0.0423
-- Rewards/margins: 0.0131
 - Rewards/rejected: -0.0554
+- Rewards/accuracies: 0.5943
+- Rewards/margins: 0.0131
+- Logps/rejected: -68.7246
+- Logps/chosen: -62.9434
+- Logits/rejected: -3.0648
+- Logits/chosen: -3.0704
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 3.0,
+    "eval_logits/chosen": -3.070441722869873,
+    "eval_logits/rejected": -3.0647687911987305,
+    "eval_logps/chosen": -62.94335174560547,
+    "eval_logps/rejected": -68.72457122802734,
+    "eval_loss": 0.6871925592422485,
+    "eval_rewards/accuracies": 0.5943308472633362,
+    "eval_rewards/chosen": -0.04231457784771919,
+    "eval_rewards/margins": 0.013129997998476028,
+    "eval_rewards/rejected": -0.05544458329677582,
+    "eval_runtime": 383.6672,
+    "eval_samples": 4304,
+    "eval_samples_per_second": 11.218,
+    "eval_steps_per_second": 1.402,
     "total_flos": 0.0,
     "train_loss": 0.000466351711199581,
     "train_runtime": 50.9431,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
   "transformers_version": "4.41.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 3.0,
+    "eval_logits/chosen": -3.070441722869873,
+    "eval_logits/rejected": -3.0647687911987305,
+    "eval_logps/chosen": -62.94335174560547,
+    "eval_logps/rejected": -68.72457122802734,
+    "eval_loss": 0.6871925592422485,
+    "eval_rewards/accuracies": 0.5943308472633362,
+    "eval_rewards/chosen": -0.04231457784771919,
+    "eval_rewards/margins": 0.013129997998476028,
+    "eval_rewards/rejected": -0.05544458329677582,
+    "eval_runtime": 383.6672,
+    "eval_samples": 4304,
+    "eval_samples_per_second": 11.218,
+    "eval_steps_per_second": 1.402
+}

runs/Jun25_14-09-14_poseidon/events.out.tfevents.1719325503.poseidon.317546.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a419ffc106c1864cd87b4cb3f581a0a5c4b74d8b85fd7d9ecdf45c6ee1b3f816
+size 841