Converting weights

README.md

@@ -1,8 +1,8 @@
 ---
-license_name: gemma-terms
-license_link: https://ai.google.dev/gemma/terms
 language:
 - en
+license_name: gemma-terms
+license_link: https://ai.google.dev/gemma/terms
 ---
 
 # LLaVA-Gemma Model Card

config.json

@@ -1,43 +1,48 @@
 {
-  "_name_or_path": "/root/.cache/huggingface/hub/models--google--gemma-2b-it/snapshots/1027d96c1638a27f01ae935cd98bac7d1a01686c/",
+  "_name_or_path": "./llava-gemma-2b-it",
   "architectures": [
-    "LlavaGemmaForCausalLM"
+    "LlavaForConditionalGeneration"
   ],
-  "attention_bias": false,
-  "attention_dropout": 0.0,
-  "bos_token_id": 2,
-  "eos_token_id": 1,
-  "freeze_mm_mlp_adapter": false,
-  "head_dim": 256,
-  "hidden_act": "gelu",
-  "hidden_size": 2048,
-  "image_aspect_ratio": "pad",
-  "initializer_range": 0.02,
-  "intermediate_size": 16384,
-  "max_position_embeddings": 8192,
-  "mm_hidden_size": 1024,
-  "mm_patch_merge_type": "flat",
-  "mm_projector_lr": null,
-  "mm_projector_type": "mlp2x_gelu",
-  "mm_use_im_patch_token": false,
-  "mm_use_im_start_end": false,
-  "mm_vision_select_feature": "patch",
-  "mm_vision_select_layer": -2,
-  "mm_vision_tower": "openai/clip-vit-large-patch14-336",
-  "model_type": "llava_gemma",
-  "num_attention_heads": 8,
-  "num_hidden_layers": 18,
-  "num_key_value_heads": 1,
-  "pad_token_id": 0,
-  "rms_norm_eps": 1e-06,
-  "rope_scaling": null,
-  "rope_theta": 10000.0,
-  "tokenizer_model_max_length": 2048,
-  "tokenizer_padding_side": "right",
-  "torch_dtype": "bfloat16",
+  "ignore_index": -100,
+  "image_token_index": 256000,
+  "model_type": "llava",
+  "projector_hidden_act": "gelu",
+  "text_config": {
+    "_name_or_path": "google/gemma-2b-it",
+    "architectures": [
+      "GemmaForCausalLM"
+    ],
+    "bos_token_id": 2,
+    "eos_token_id": 1,
+    "head_dim": 256,
+    "hidden_act": "gelu",
+    "hidden_size": 2048,
+    "intermediate_size": 16384,
+    "max_position_embeddings": 8192,
+    "model_type": "gemma",
+    "num_attention_heads": 8,
+    "num_hidden_layers": 18,
+    "num_key_value_heads": 1,
+    "pad_token_id": 0,
+    "rope_scaling": null,
+    "tie_word_embeddings": true,
+    "torch_dtype": "bfloat16",
+    "vocab_size": 256064
+  },
+  "torch_dtype": "float32",
   "transformers_version": "4.39.0.dev0",
-  "tune_mm_mlp_adapter": false,
-  "use_cache": false,
-  "use_mm_proj": true,
-  "vocab_size": 256000
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vocab_size": 256064
 }

generation_config.json

@@ -1,19 +1,7 @@
 {
   "_from_model_config": true,
-  "attn_softmax_bf16": null,
   "bos_token_id": 2,
-  "bucket_internal": null,
-  "bucket_size": -1,
   "eos_token_id": 1,
-  "flash_attention_recompute": null,
-  "ignore_eos": null,
-  "kv_cache_fp8": null,
-  "limit_hpu_graphs": null,
   "pad_token_id": 0,
-  "reduce_recompile": null,
-  "reuse_cache": null,
-  "static_shapes": null,
-  "transformers_version": "4.39.0.dev0",
-  "trim_logits": null,
-  "use_flash_attention": null
+  "transformers_version": "4.39.0.dev0"
 }

model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a34f002270e12d50917ff5d307951ed107140530372c2d549aea081e327b4f6f
+size 4964385864

model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21d7fe224a3dced4a95cd2316b9d131a5c80705115849dd980eafbfe108d5e5c
+size 4999820616

model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40c198b0d6b4170bd3d8eb708a66a672b76ae70eece7ab843b8ed3605d97c85c
+size 1300294808