Update

README.md

@@ -9,13 +9,13 @@ license: apache-2.0
   <img src="./icon.png" alt="Logo" width="350">
 </p>
 
-📖 [Technical report](https://arxiv.org/abs/2402.11530) | 🏠 [Code](https://github.com/BAAI-DCAI/Bunny) | 🐰 [3B Demo](https://wisemodel.cn/spaces/baai/Bunny) | 🐰 [8B Demo](https://252412006bcde38bfa.gradio.live/)
+📖 [Technical report](https://arxiv.org/abs/2402.11530) | 🏠 [Code](https://github.com/BAAI-DCAI/Bunny) | 🐰 [3B Demo](https://wisemodel.cn/spaces/baai/Bunny) | 🐰 [8B Demo](https://2e09fec5116a0ba343.gradio.live)
 
 This is Bunny-Llama-3-8B-V.
 
-Bunny is a family of lightweight but powerful multimodal models. It offers multiple plug-and-play vision encoders, like EVA-CLIP, SigLIP and language backbones, including Llama-3-8B, Phi-1.5, StableLM-2 and Phi-2. To compensate for the decrease in model size, we construct more informative training data by curated selection from a broader data source.
+Bunny is a family of lightweight but powerful multimodal models. It offers multiple plug-and-play vision encoders, like EVA-CLIP, SigLIP and language backbones, including Llama-3-8B, Phi-1.5, StableLM-2, Qwen1.5, MiniCPM and Phi-2. To compensate for the decrease in model size, we construct more informative training data by curated selection from a broader data source.
 
-We provide Bunny-Llama-3-8B-V, which is built upon [SigLIP](https://huggingface.co/google/siglip-so400m-patch14-384) and [Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B). More details about this model can be found in [GitHub](https://github.com/BAAI-DCAI/Bunny).
+We provide Bunny-Llama-3-8B-V, which is built upon [SigLIP](https://huggingface.co/google/siglip-so400m-patch14-384) and [Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct). More details about this model can be found in [GitHub](https://github.com/BAAI-DCAI/Bunny).
 
 ![comparison](comparison.png)
 

config.json

@@ -39,5 +39,5 @@
   "unfreeze_vision_tower": true,
   "use_cache": true,
   "use_mm_proj": true,
-  "vocab_size": 128257
+  "vocab_size": 128256
 }

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ec0a7fd9ad460c3e0a4531b16c290472d8d133c348e9649efa2e9458e936a39
-size 4976706784
+oid sha256:8b002c3d2062a306cb9317fb6c7d3776d764cee429f6871c97b7d10c474340fd
+size 4976698592

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ec257e1403e02aaac7d2d79e67ad4f5378c4b041037aa00862b1983b014729b1
+oid sha256:c6fbc96f0928d48703340ad0c75639aa2c4468d3908e9da197babcdb27b0b483
 size 4999802616

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb70b23deec1b9d9512676ba5c42ea0d3c980c1cba387fc9cb14e46be6153b08
+oid sha256:d98fe9be8120810794a907796857179ae66e3d94a621c66ced5728e3bfbed018
 size 4915916080

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:423517f46806e703f787ecb8ac332c729f167baeaca7e500da9462583ed8cda1
-size 2067676408
+oid sha256:5aa5017356655dc29634c195c2ca2f515f3f0e1ba4aeac0454af8d946a520744
+size 2067668216

model.safetensors.index.json

@@ -1,6 +1,6 @@
 {
   "metadata": {
-    "total_size": 16959998080
+    "total_size": 16959981696
   },
   "weight_map": {
     "lm_head.weight": "model-00004-of-00004.safetensors",

special_tokens_map.json

@@ -12,6 +12,5 @@
     "normalized": false,
     "rstrip": false,
     "single_word": false
-  },
-  "unk_token": "<unk>"
+  }
 }

tokenizer_config.json

@@ -2050,6 +2050,7 @@
     }
   },
   "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
   "clean_up_tokenization_spaces": true,
   "eos_token": "<|end_of_text|>",
   "model_input_names": [