Model save

README.md

@@ -1,15 +1,11 @@
 ---
 base_model: dmis-lab/selfbiorag_7b
 tags:
-- alignment-handbook
-- trl
-- sft
-- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- HuggingFaceH4/deita-10k-v0-sft
+- generator
 model-index:
 - name: selfbiorag-7b-wo-live_qa-iter-sft-step1
   results: []
@@ -20,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # selfbiorag-7b-wo-live_qa-iter-sft-step1
 
-This model is a fine-tuned version of [dmis-lab/selfbiorag_7b](https://huggingface.co/dmis-lab/selfbiorag_7b) on the HuggingFaceH4/deita-10k-v0-sft dataset.
+This model is a fine-tuned version of [dmis-lab/selfbiorag_7b](https://huggingface.co/dmis-lab/selfbiorag_7b) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.9657
+- Loss: 1.1582
 
 ## Model description
 
@@ -59,14 +55,14 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.5018        | 1.0   | 5    | 1.7469          |
-| 1.1848        | 2.0   | 10   | 1.9052          |
-| 0.9623        | 3.0   | 15   | 1.9657          |
+| 1.2659        | 0.91  | 5    | 1.2054          |
+| 1.0094        | 2.0   | 11   | 1.1690          |
+| 0.8249        | 2.73  | 15   | 1.1582          |
 
 
 ### Framework versions
 
-- Transformers 4.39.0.dev0
-- Pytorch 2.1.2
+- Transformers 4.38.2
+- Pytorch 2.1.2+cu121
 - Datasets 2.14.6
 - Tokenizers 0.15.2

all_results.json

@@ -1,13 +1,8 @@
 {
-    "epoch": 3.0,
-    "eval_loss": 1.9656569957733154,
-    "eval_runtime": 1.8218,
-    "eval_samples": 100,
-    "eval_samples_per_second": 4.391,
-    "eval_steps_per_second": 0.549,
-    "train_loss": 1.22816797097524,
-    "train_runtime": 294.6904,
+    "epoch": 2.73,
+    "train_loss": 1.0476287603378296,
+    "train_runtime": 318.5273,
     "train_samples": 4848,
-    "train_samples_per_second": 3.176,
-    "train_steps_per_second": 0.051
+    "train_samples_per_second": 3.306,
+    "train_steps_per_second": 0.047
 }

config.json

@@ -22,7 +22,7 @@
   "rope_theta": 10000.0,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.39.0.dev0",
-  "use_cache": true,
+  "transformers_version": "4.38.2",
+  "use_cache": false,
   "vocab_size": 32016
 }

generation_config.json

@@ -6,5 +6,5 @@
   "pad_token_id": 0,
   "temperature": 0.6,
   "top_p": 0.9,
-  "transformers_version": "4.39.0.dev0"
+  "transformers_version": "4.38.2"
 }

model-00001-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2e508574c3f3efaf638d758870e4e73ee29de4828cbf4b295fd2cc366e85dfd
+oid sha256:93c4c2b9b3ec5dc0efd95f4b203ce8723fffa0e89d2903299b3f2c5696b4c420
 size 4939116424

model-00002-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb1b3c56456213b69e50d195f19bcd64b29abe831b726c28e7df2bccd6891081
+oid sha256:5b277ce0ae848f0124609ae2982d0558a2c8015cab2bad35611540a3ef75a72e
 size 4947390880

model-00003-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9220410beb9d45a759a152c57e89d5b510fdc214aac830a65438c92f7e2bde35
+oid sha256:20d276b07d457aaa761e20c71e1d7881afa6a139da4bbe78cbc3527d5e57b3b0
 size 3590619888

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "train_loss": 1.22816797097524,
-    "train_runtime": 294.6904,
+    "epoch": 2.73,
+    "train_loss": 1.0476287603378296,
+    "train_runtime": 318.5273,
     "train_samples": 4848,
-    "train_samples_per_second": 3.176,
-    "train_steps_per_second": 0.051
+    "train_samples_per_second": 3.306,
+    "train_steps_per_second": 0.047
 }

trainer_state.json

@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0,
+  "epoch": 2.7272727272727275,
   "eval_steps": 500,
   "global_step": 15,
   "is_hyper_param_search": false,
@@ -9,65 +9,65 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.2,
-      "grad_norm": 8.297603900150138,
+      "epoch": 0.18,
+      "grad_norm": 8.112652061034474,
       "learning_rate": 1e-05,
-      "loss": 1.6795,
+      "loss": 1.4798,
       "step": 1
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 2.7714905692865806,
+      "epoch": 0.91,
+      "grad_norm": 2.848398983756101,
       "learning_rate": 1.7485107481711014e-05,
-      "loss": 1.5018,
+      "loss": 1.2659,
       "step": 5
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 1.746938943862915,
-      "eval_runtime": 1.6252,
-      "eval_samples_per_second": 4.922,
-      "eval_steps_per_second": 0.615,
+      "epoch": 0.91,
+      "eval_loss": 1.2053821086883545,
+      "eval_runtime": 1.8264,
+      "eval_samples_per_second": 3.833,
+      "eval_steps_per_second": 0.548,
       "step": 5
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 2.355589361090063,
+      "epoch": 1.82,
+      "grad_norm": 2.6874796106077996,
       "learning_rate": 6.453951129574644e-06,
-      "loss": 1.1848,
+      "loss": 1.0094,
       "step": 10
     },
     {
       "epoch": 2.0,
-      "eval_loss": 1.9052479267120361,
-      "eval_runtime": 1.5781,
-      "eval_samples_per_second": 5.069,
-      "eval_steps_per_second": 0.634,
-      "step": 10
+      "eval_loss": 1.1690208911895752,
+      "eval_runtime": 1.8589,
+      "eval_samples_per_second": 3.766,
+      "eval_steps_per_second": 0.538,
+      "step": 11
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 1.8474736569347063,
+      "epoch": 2.73,
+      "grad_norm": 1.7499037267304558,
       "learning_rate": 0.0,
-      "loss": 0.9623,
+      "loss": 0.8249,
       "step": 15
     },
     {
-      "epoch": 3.0,
-      "eval_loss": 1.9656569957733154,
-      "eval_runtime": 1.5681,
-      "eval_samples_per_second": 5.102,
-      "eval_steps_per_second": 0.638,
+      "epoch": 2.73,
+      "eval_loss": 1.1581600904464722,
+      "eval_runtime": 1.8666,
+      "eval_samples_per_second": 3.75,
+      "eval_steps_per_second": 0.536,
       "step": 15
     },
     {
-      "epoch": 3.0,
+      "epoch": 2.73,
       "step": 15,
       "total_flos": 3088349921280.0,
-      "train_loss": 1.22816797097524,
-      "train_runtime": 294.6904,
-      "train_samples_per_second": 3.176,
-      "train_steps_per_second": 0.051
+      "train_loss": 1.0476287603378296,
+      "train_runtime": 318.5273,
+      "train_samples_per_second": 3.306,
+      "train_steps_per_second": 0.047
     }
   ],
   "logging_steps": 5,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:91e3f59bd6deed88623a8848dda32804386130d521096cbe9354c14e9906dfd5
+oid sha256:f4860520a8292e2b6575761ceba29cc22d6e44d02462c96fceff35960661933a
 size 6200