Upload folder using huggingface_hub

added_tokens.json

@@ -0,0 +1,6 @@
+{
+  "<PAD>": 32003,
+  "<answer>": 32002,
+  "<image>": 32001,
+  "<|endofchunk|>": 32000
+}

config.json

@@ -0,0 +1,173 @@
+{
+  "_commit_hash": null,
+  "_name_or_path": "",
+  "architectures": [
+    "FlamingoForConditionalGeneration"
+  ],
+  "cross_attn_every_n_layers": 4,
+  "feature_as_input": true,
+  "hidden_size": 4096,
+  "model_type": "flamingo",
+  "only_attend_immediate_media": true,
+  "only_attend_previous": true,
+  "text_config": {
+    "_name_or_path": "luodian/llama-7b-hf",
+    "add_cross_attention": false,
+    "architectures": null,
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": 1,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": 2,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "silu",
+    "hidden_size": 4096,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 11008,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "max_position_embeddings": 2048,
+    "min_length": 0,
+    "model_type": "llama",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 32,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_hidden_layers": 32,
+    "num_key_value_heads": 32,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": 0,
+    "prefix": null,
+    "pretraining_tp": 1,
+    "problem_type": null,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "rms_norm_eps": 1e-06,
+    "rope_scaling": null,
+    "rope_theta": 10000.0,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": false,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": null,
+    "torchscript": false,
+    "transformers_version": "4.37.2",
+    "typical_p": 1.0,
+    "use_bfloat16": false,
+    "use_cache": true,
+    "vocab_size": 32004
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": null,
+  "vision_config": {
+    "_name_or_path": "openai/clip-vit-large-patch14",
+    "add_cross_attention": false,
+    "architectures": null,
+    "attention_dropout": 0.0,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "quick_gelu",
+    "hidden_size": 1024,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "image_size": 224,
+    "initializer_factor": 1.0,
+    "initializer_range": 0.02,
+    "intermediate_size": 4096,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "layer_norm_eps": 1e-05,
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "min_length": 0,
+    "model_type": "clip_vision_model",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 16,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_channels": 3,
+    "num_hidden_layers": 24,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": null,
+    "patch_size": 14,
+    "prefix": null,
+    "problem_type": null,
+    "projection_dim": 512,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": null,
+    "torchscript": false,
+    "transformers_version": "4.37.2",
+    "typical_p": 1.0,
+    "use_bfloat16": false
+  }
+}

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98214b2ae6d47f33345278ac36f8ea1b1a3ff1c9ec6f3a5071889be90cf5d274
+size 4968434376

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd86e7463b188a271e3f53c272f4f7af0b0cbf0b371e220d0b360aa630a4e6d3
+size 4918104252

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f8fc1a1593396c1a79c387b7412093160f6e4cfe7bf76eca33cd7c1af126cba
+size 4886663364

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5ba31f41938c49457f3745e404808d92064471682b652b66d4498b2e7e301a6
+size 1061697000

model.safetensors.index.json

@@ -0,0 +1,488 @@
+{
+  "metadata": {
+    "total_size": 15834832928
+  },
+  "weight_map": {
+    "lang_encoder.lm_head.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.embed_tokens.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.0.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.1.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.10.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.attn.norm.bias": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.attn.norm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.attn.to_kv.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.attn.to_out.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.attn.to_q.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.attn_gate": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.feed_forward.0.bias": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.feed_forward.0.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.feed_forward.1.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.feed_forward.3.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.11.gated_cross_attn_layer.ff_gate": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.12.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.13.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.14.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.attn.norm.bias": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.attn.norm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.attn.to_kv.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.attn.to_out.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.attn.to_q.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.attn_gate": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.feed_forward.0.bias": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.feed_forward.0.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.feed_forward.1.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.feed_forward.3.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.15.gated_cross_attn_layer.ff_gate": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.16.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.17.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.18.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.attn.norm.bias": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.attn.norm.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.attn.to_kv.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.attn.to_out.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.attn.to_q.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.attn_gate": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.feed_forward.0.bias": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.feed_forward.0.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.feed_forward.1.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.feed_forward.3.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.19.gated_cross_attn_layer.ff_gate": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.2.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.20.decoder_layer.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.21.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.22.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.attn.norm.bias": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.attn.norm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.attn.to_kv.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.attn.to_out.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.attn.to_q.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.attn_gate": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.feed_forward.0.bias": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.feed_forward.0.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.feed_forward.1.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.feed_forward.3.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.23.gated_cross_attn_layer.ff_gate": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.24.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.25.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.26.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.attn.norm.bias": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.attn.norm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.attn.to_kv.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.attn.to_out.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.attn.to_q.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.attn_gate": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.feed_forward.0.bias": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.feed_forward.0.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.feed_forward.1.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.feed_forward.3.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.27.gated_cross_attn_layer.ff_gate": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.28.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.29.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.attn.norm.bias": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.attn.norm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.attn.to_kv.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.attn.to_out.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.attn.to_q.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.attn_gate": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.feed_forward.0.bias": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.feed_forward.0.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.feed_forward.1.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.feed_forward.3.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.3.gated_cross_attn_layer.ff_gate": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.30.decoder_layer.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.input_layernorm.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.decoder_layer.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.attn.norm.bias": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.attn.norm.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.attn.to_kv.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.attn.to_out.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.attn.to_q.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.attn_gate": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.feed_forward.0.bias": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.feed_forward.0.weight": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.feed_forward.1.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.feed_forward.3.weight": "model-00004-of-00004.safetensors",
+    "lang_encoder.model.layers.31.gated_cross_attn_layer.ff_gate": "model-00003-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.4.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.5.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.6.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.attn.norm.bias": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.attn.norm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.attn.to_kv.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.attn.to_out.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.attn.to_q.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.attn_gate": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.feed_forward.0.bias": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.feed_forward.0.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.feed_forward.1.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.feed_forward.3.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.7.gated_cross_attn_layer.ff_gate": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.8.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.layers.9.decoder_layer.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "lang_encoder.model.norm.weight": "model-00004-of-00004.safetensors",
+    "perceiver.frame_embs": "model-00004-of-00004.safetensors",
+    "perceiver.latents": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.feed_forward.0.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.feed_forward.0.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.feed_forward.1.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.feed_forward.3.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.norm_latents.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.norm_latents.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.norm_media.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.norm_media.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.to_kv.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.to_out.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.0.to_q.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.feed_forward.0.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.feed_forward.0.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.feed_forward.1.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.feed_forward.3.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.norm_latents.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.norm_latents.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.norm_media.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.norm_media.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.to_kv.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.to_out.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.1.to_q.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.feed_forward.0.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.feed_forward.0.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.feed_forward.1.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.feed_forward.3.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.norm_latents.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.norm_latents.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.norm_media.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.norm_media.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.to_kv.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.to_out.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.2.to_q.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.feed_forward.0.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.feed_forward.0.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.feed_forward.1.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.feed_forward.3.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.norm_latents.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.norm_latents.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.norm_media.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.norm_media.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.to_kv.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.to_out.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.3.to_q.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.feed_forward.0.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.feed_forward.0.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.feed_forward.1.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.feed_forward.3.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.norm_latents.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.norm_latents.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.norm_media.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.norm_media.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.to_kv.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.to_out.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.4.to_q.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.feed_forward.0.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.feed_forward.0.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.feed_forward.1.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.feed_forward.3.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.norm_latents.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.norm_latents.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.norm_media.bias": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.norm_media.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.to_kv.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.to_out.weight": "model-00004-of-00004.safetensors",
+    "perceiver.layers.5.to_q.weight": "model-00004-of-00004.safetensors",
+    "perceiver.norm.bias": "model-00004-of-00004.safetensors",
+    "perceiver.norm.weight": "model-00004-of-00004.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,53 @@
+{
+  "additional_special_tokens": [
+    {
+      "content": "<|endofchunk|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "<image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "<answer>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    }
+  ],
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<PAD>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

tokenizer_config.json

@@ -0,0 +1,78 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32000": {
+      "content": "<|endofchunk|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32001": {
+      "content": "<image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32002": {
+      "content": "<answer>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32003": {
+      "content": "<PAD>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endofchunk|>",
+    "<image>",
+    "<answer>"
+  ],
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<PAD>",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}