End of training

README.md

@@ -33,14 +33,14 @@ More information needed
 The following hyperparameters were used during training:
 - learning_rate: 2e-05
 - train_batch_size: 2
-- eval_batch_size: 8
+- eval_batch_size: 4
 - seed: 42
-- gradient_accumulation_steps: 16
-- total_train_batch_size: 32
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 16
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 3
+- num_epochs: 4
 
 ### Training results
 

generation_config.json

@@ -6,8 +6,7 @@
     151643
   ],
   "pad_token_id": 151643,
-  "temperature": 0.01,
-  "top_k": 1,
-  "top_p": 0.001,
+  "temperature": 0.8,
+  "top_p": 0.95,
   "transformers_version": "4.46.0"
 }

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8570fe809f6f5302ccc34ac30b0b60c0b1e6fceaff8024e1cbf578b8d9d470c0
+oid sha256:10e7c7cebc1a39694902f8b478d0f8d9081130b5397aacd1e568b6fbf9b3f01c
 size 4966659944

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ace3ecdfd1c1f14e3144fcd0532232fab55612ba0021c0eeeee1deb4ab18692
+oid sha256:38bb9e57c8e07d5d1750f39bdc822e9d45460737b3d9b35cf9640560604b29f8
 size 4991495816

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d1347e424d39dfb1f056fa4bff9bd70c0e47b2af80834b3edf1af263eb05213e
+oid sha256:a38e53ab9b1f5f501c40160d89d8b407e14771fa89354ab4ee0f4bd7bf7b2740
 size 4932751040

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:15695fc010c5ee4ed457200fc59d12e1e8fd78381fb849dcf3edc121da45e1e1
+oid sha256:fabc46bfd3157ce05711647ddbba2d86082f304bfa38605e1397745b46ba6bcf
 size 1691924384

tokenizer_config.json

@@ -137,8 +137,8 @@
   "model_max_length": 32768,
   "pad_token": "<|endoftext|>",
   "padding_side": "right",
-  "processor_class": "Qwen2VLProcessor",
   "split_special_tokens": false,
   "tokenizer_class": "Qwen2Tokenizer",
-  "unk_token": null
+  "unk_token": null,
+  "use_fast": true
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3ac3992e6743d379f816b8cd04360186c72fde6fc27b5f08d6b1330ce94b288f
-size 5304
+oid sha256:d231ba54507ce4fa6a6b4f166310e855afbecbd95170a261d41b7d4031caab89
+size 5240