End of training

README.md

@@ -42,7 +42,7 @@ The following hyperparameters were used during training:
 - total_train_batch_size: 16
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 30
+- num_epochs: 100
 - mixed_precision_training: Native AMP
 
 ### Training results

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd0ef8f05ca17373d9e41472a79cfd9ec58f3c85e1448791e821ff1de43d4663
+oid sha256:031e6e034983ae6fc9e5c8b5a7eca83192008c78ed6d3ef5562e703845a2f00c
 size 756399848

tokenizer_config.json

@@ -44,7 +44,7 @@
     }
   },
   "bos_token": "<s>",
-  "chat_template": "A curious user asks an AI about CAD models based on provided images. The AI provides a detailed and polite answer identifying the model. For each dataset entry:\nUSER: What CAD model do these images represent? <image>\nASSISTANT: The CAD model these images represent is {{ answer }}. {{eos_token}}",
+  "chat_template": "A curious user asks an AI about CAD models based on provided images. The AI provides a detailed and polite answer identifying the model. For each dataset entry:\nUSER: The image you see is a combination of different views of a CAD model. What CAD model do these images represent? \n<image>\nASSISTANT: The CAD model these images represent is: {{ answer }}. {{eos_token}}",
   "clean_up_tokenization_spaces": false,
   "eos_token": "</s>",
   "legacy": false,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:19c1025e1f5686fec19bb23c416cb1cc18eb7ac87cd667101f99e48f47630cfb
+oid sha256:fa384375d6f91e1661015c8f88755966751845b7386b2a8e577d36eaea406859
 size 4984