Training in progress, step 30

config.json

@@ -8,11 +8,9 @@
   "model_type": "vision-text-dual-encoder",
   "projection_dim": 512,
   "text_config": {
-    "_name_or_path": "roberta-base",
+    "_name_or_path": "allenai/biomed_roberta_base",
     "add_cross_attention": false,
-    "architectures": [
-      "RobertaForMaskedLM"
-    ],
+    "architectures": null,
     "attention_probs_dropout_prob": 0.1,
     "bad_words_ids": null,
     "begin_suppress_tokens": null,
@@ -90,7 +88,7 @@
   "torch_dtype": "float32",
   "transformers_version": "4.41.2",
   "vision_config": {
-    "_name_or_path": "openai/clip-vit-base-patch32",
+    "_name_or_path": "openai/clip-vit-large-patch14-336",
     "add_cross_attention": false,
     "architectures": null,
     "attention_dropout": 0.0,
@@ -111,15 +109,15 @@
     "forced_bos_token_id": null,
     "forced_eos_token_id": null,
     "hidden_act": "quick_gelu",
-    "hidden_size": 768,
+    "hidden_size": 1024,
     "id2label": {
       "0": "LABEL_0",
       "1": "LABEL_1"
     },
-    "image_size": 224,
+    "image_size": 336,
     "initializer_factor": 1.0,
     "initializer_range": 0.02,
-    "intermediate_size": 3072,
+    "intermediate_size": 4096,
     "is_decoder": false,
     "is_encoder_decoder": false,
     "label2id": {
@@ -132,20 +130,20 @@
     "min_length": 0,
     "model_type": "clip_vision_model",
     "no_repeat_ngram_size": 0,
-    "num_attention_heads": 12,
+    "num_attention_heads": 16,
     "num_beam_groups": 1,
     "num_beams": 1,
     "num_channels": 3,
-    "num_hidden_layers": 12,
+    "num_hidden_layers": 24,
     "num_return_sequences": 1,
     "output_attentions": false,
     "output_hidden_states": false,
     "output_scores": false,
     "pad_token_id": null,
-    "patch_size": 32,
+    "patch_size": 14,
     "prefix": null,
     "problem_type": null,
-    "projection_dim": 512,
+    "projection_dim": 768,
     "pruned_heads": {},
     "remove_invalid_values": false,
     "repetition_penalty": 1.0,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa1cfee1a3f3a86f86c4ccb5f265543a870a49cead43e0806e7e141372f3f4ed
-size 851603588
+oid sha256:ec27fd6b313dfe5163a2352de293480869155443fc2fd2e76aac55f2a7dc2b15
+size 1716360444

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea249cd364526eab937ad6487a7f81bc8fe5346a1416e3119d05e564ae6f62c3
+oid sha256:43d91b1496178ac406352964cea26801514f7d8531618943d63c80852c83fa37
 size 5112