Model save

README.md

@@ -2,13 +2,9 @@
 license: other
 library_name: peft
 tags:
-- alignment-handbook
-- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
-datasets:
-- ruozhiba
 base_model: 01-ai/Yi-6B
 model-index:
 - name: Yi-6B-ruozhiba-5e-4-50
@@ -20,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Yi-6B-ruozhiba-5e-4-50
 
-This model is a fine-tuned version of [01-ai/Yi-6B](https://huggingface.co/01-ai/Yi-6B) on the ruozhiba dataset.
+This model is a fine-tuned version of [01-ai/Yi-6B](https://huggingface.co/01-ai/Yi-6B) on the None dataset.
 It achieves the following results on the evaluation set:
 - Loss: 3.5394
 
@@ -48,7 +44,7 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 20
+- num_epochs: 10
 
 ### Training results
 
@@ -73,7 +69,6 @@ The following hyperparameters were used during training:
 | 0.0395        | 17.0  | 935  | 3.5320          |
 | 0.0408        | 18.0  | 990  | 3.5371          |
 | 0.0468        | 19.0  | 1045 | 3.5391          |
-| 0.0358        | 20.0  | 1100 | 3.5394          |
 
 
 ### Framework versions

adapter_config.json

@@ -19,13 +19,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "down_proj",
-    "k_proj",
     "gate_proj",
-    "v_proj",
+    "k_proj",
     "o_proj",
-    "up_proj"
+    "v_proj",
+    "down_proj",
+    "up_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM"
 }

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 20.0,
     "eval_loss": 3.5393826961517334,
-    "eval_runtime": 4.9183,
+    "eval_runtime": 6.2217,
     "eval_samples": 23,
-    "eval_samples_per_second": 4.676,
-    "eval_steps_per_second": 1.22,
-    "train_loss": 0.31520286625081845,
-    "train_runtime": 10333.7168,
+    "eval_samples_per_second": 3.697,
+    "eval_steps_per_second": 0.964,
+    "train_loss": 0.0,
+    "train_runtime": 10.4648,
     "train_samples": 217,
-    "train_samples_per_second": 0.42,
-    "train_steps_per_second": 0.106
+    "train_samples_per_second": 207.361,
+    "train_steps_per_second": 52.557
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
     "eval_loss": 3.5393826961517334,
-    "eval_runtime": 4.9183,
+    "eval_runtime": 6.2217,
     "eval_samples": 23,
-    "eval_samples_per_second": 4.676,
-    "eval_steps_per_second": 1.22
+    "eval_samples_per_second": 3.697,
+    "eval_steps_per_second": 0.964
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.31520286625081845,
-    "train_runtime": 10333.7168,
+    "train_loss": 0.0,
+    "train_runtime": 10.4648,
     "train_samples": 217,
-    "train_samples_per_second": 0.42,
-    "train_steps_per_second": 0.106
+    "train_samples_per_second": 207.361,
+    "train_steps_per_second": 52.557
 }

trainer_state.json

@@ -2006,38 +2006,20 @@
       "loss": 0.0358,
       "step": 1100
     },
-    {
-      "epoch": 20.0,
-      "gpt4_scores": 0.58,
-      "step": 1100
-    },
-    {
-      "epoch": 20.0,
-      "std": 0.12692517480783708,
-      "step": 1100
-    },
-    {
-      "epoch": 20.0,
-      "eval_loss": 3.5393826961517334,
-      "eval_runtime": 4.9612,
-      "eval_samples_per_second": 4.636,
-      "eval_steps_per_second": 1.209,
-      "step": 1100
-    },
     {
       "epoch": 20.0,
       "step": 1100,
       "total_flos": 3.792163606268314e+16,
-      "train_loss": 0.31520286625081845,
-      "train_runtime": 10333.7168,
-      "train_samples_per_second": 0.42,
-      "train_steps_per_second": 0.106
+      "train_loss": 0.0,
+      "train_runtime": 10.4648,
+      "train_samples_per_second": 207.361,
+      "train_steps_per_second": 52.557
     }
   ],
   "logging_steps": 4,
-  "max_steps": 1100,
+  "max_steps": 550,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 20,
+  "num_train_epochs": 10,
   "save_steps": 55,
   "total_flos": 3.792163606268314e+16,
   "train_batch_size": 4,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d43b9ca7dbac23c8c7424d5fd80ec777548309c8c41f65dd80f1563a62bd651
+oid sha256:067278813e5531ea0ebcd32f207e9c4e28e7ff60c4fe15536189c80bdc3266ee
 size 4792