update the checkpoints of PLM and Qwen after training on PanoVQA

plm/plm3b/adaptor_config.json

@@ -0,0 +1,7 @@
+{
+    "adaptor_name": null,
+    "bottle_dim": 512,
+    "sparse_k": 128,
+    "index_heads": 4,
+    "index_dim": 16
+}

plm/plm3b/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

plm/plm3b/checkpoint-100/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

plm/plm3b/checkpoint-100/config.json

@@ -0,0 +1,65 @@
+{
+  "architectures": [
+    "Qwen2_5_VLForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_token_id": 151655,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 128000,
+  "max_window_layers": 70,
+  "model_type": "qwen2_5_vl",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": {
+    "mrope_section": [
+      16,
+      24,
+      24
+    ],
+    "rope_type": "default",
+    "type": "default"
+  },
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.50.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "video_token_id": 151656,
+  "vision_config": {
+    "depth": 32,
+    "fullatt_block_indexes": [
+      7,
+      15,
+      23,
+      31
+    ],
+    "hidden_act": "silu",
+    "hidden_size": 1280,
+    "in_channels": 3,
+    "in_chans": 3,
+    "intermediate_size": 3420,
+    "model_type": "qwen2_5_vl",
+    "num_heads": 16,
+    "out_hidden_size": 2048,
+    "patch_size": 14,
+    "spatial_merge_size": 2,
+    "spatial_patch_size": 14,
+    "temporal_patch_size": 2,
+    "tokens_per_second": 2,
+    "torch_dtype": "bfloat16",
+    "window_size": 112
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652,
+  "vision_token_id": 151654,
+  "vocab_size": 151936
+}

plm/plm3b/checkpoint-100/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "attn_implementation": "flash_attention_2",
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 1e-06,
+  "transformers_version": "4.50.0"
+}

plm/plm3b/checkpoint-100/global_step100/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a720a82be02568beee7d3e270378af66fd515b745faaf7f728e8a5a4437d747
+size 37471776486

plm/plm3b/checkpoint-100/global_step100/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80904d25bb55d410ac997b13e922cfa6e97858f7a5fd84129ecedb5d1e2c9074
+size 1264465844

plm/plm3b/checkpoint-100/latest

@@ -0,0 +1 @@
+global_step100

plm/plm3b/checkpoint-100/pytorch_model-00001-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e75e1a4aa52d25cdc125c54b5186fb02f205319e660c535fca84a76ad0a0d1f
+size 4997892186

plm/plm3b/checkpoint-100/pytorch_model-00002-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdfd5ca43564d82ce6fc640bb503d4641f1429f727307a27c1a476991a1b4d47
+size 2511631122

plm/plm3b/checkpoint-100/pytorch_model.bin.index.json

@@ -0,0 +1,832 @@
+{
+  "metadata": {
+    "total_size": 7509245952
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00001-of-00002.bin",
+    "model.embed_tokens.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.20.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.30.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.norm.weight": "pytorch_model-00002-of-00002.bin",
+    "visual.blocks.0.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.ln_q.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.0.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.0.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.2.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.patch_embed.proj.weight": "pytorch_model-00001-of-00002.bin"
+  }
+}

plm/plm3b/checkpoint-100/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:787db515f7b6743e3b35596e1138c851cf5634dd7a63c24ff9474f6ad1af981c
+size 14244

plm/plm3b/checkpoint-100/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4caa274ad45836431e476763cb2769c09de1a4149c1b64047932813499088bcb
+size 1064

plm/plm3b/checkpoint-100/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

plm/plm3b/checkpoint-100/tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

plm/plm3b/checkpoint-100/trainer_state.json

@@ -0,0 +1,734 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.36764705882352944,
+  "eval_steps": 500,
+  "global_step": 100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003676470588235294,
+      "grad_norm": 13.434047998912797,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 2.3324,
+      "step": 1
+    },
+    {
+      "epoch": 0.007352941176470588,
+      "grad_norm": 12.845819001528637,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 2.3146,
+      "step": 2
+    },
+    {
+      "epoch": 0.011029411764705883,
+      "grad_norm": 13.642182825241747,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 2.3436,
+      "step": 3
+    },
+    {
+      "epoch": 0.014705882352941176,
+      "grad_norm": 13.177595637373734,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 2.1726,
+      "step": 4
+    },
+    {
+      "epoch": 0.01838235294117647,
+      "grad_norm": 12.647046172843465,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 2.2284,
+      "step": 5
+    },
+    {
+      "epoch": 0.022058823529411766,
+      "grad_norm": 12.96683330506239,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.251,
+      "step": 6
+    },
+    {
+      "epoch": 0.025735294117647058,
+      "grad_norm": 12.856658525760958,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 2.0926,
+      "step": 7
+    },
+    {
+      "epoch": 0.029411764705882353,
+      "grad_norm": 10.167587096031813,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 2.074,
+      "step": 8
+    },
+    {
+      "epoch": 0.03308823529411765,
+      "grad_norm": 12.228521156970206,
+      "learning_rate": 5e-06,
+      "loss": 1.986,
+      "step": 9
+    },
+    {
+      "epoch": 0.03676470588235294,
+      "grad_norm": 9.687971167815158,
+      "learning_rate": 4.999821641793195e-06,
+      "loss": 2.0738,
+      "step": 10
+    },
+    {
+      "epoch": 0.04044117647058824,
+      "grad_norm": 9.005845290500895,
+      "learning_rate": 4.999286592622096e-06,
+      "loss": 1.7819,
+      "step": 11
+    },
+    {
+      "epoch": 0.04411764705882353,
+      "grad_norm": 7.142208208166851,
+      "learning_rate": 4.998394928831034e-06,
+      "loss": 1.5259,
+      "step": 12
+    },
+    {
+      "epoch": 0.04779411764705882,
+      "grad_norm": 6.173050515660737,
+      "learning_rate": 4.9971467776484526e-06,
+      "loss": 1.5163,
+      "step": 13
+    },
+    {
+      "epoch": 0.051470588235294115,
+      "grad_norm": 6.494634601590806,
+      "learning_rate": 4.995542317168756e-06,
+      "loss": 1.4624,
+      "step": 14
+    },
+    {
+      "epoch": 0.05514705882352941,
+      "grad_norm": 6.170011124682238,
+      "learning_rate": 4.993581776326901e-06,
+      "loss": 1.6327,
+      "step": 15
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 5.756663516704019,
+      "learning_rate": 4.991265434865726e-06,
+      "loss": 1.4856,
+      "step": 16
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 6.442210901739369,
+      "learning_rate": 4.988593623296038e-06,
+      "loss": 1.5692,
+      "step": 17
+    },
+    {
+      "epoch": 0.0661764705882353,
+      "grad_norm": 7.493967445703748,
+      "learning_rate": 4.985566722849454e-06,
+      "loss": 1.6112,
+      "step": 18
+    },
+    {
+      "epoch": 0.06985294117647059,
+      "grad_norm": 6.885056202242848,
+      "learning_rate": 4.9821851654240025e-06,
+      "loss": 1.5994,
+      "step": 19
+    },
+    {
+      "epoch": 0.07352941176470588,
+      "grad_norm": 6.019009530543658,
+      "learning_rate": 4.9784494335225e-06,
+      "loss": 1.4061,
+      "step": 20
+    },
+    {
+      "epoch": 0.07720588235294118,
+      "grad_norm": 7.03669923794518,
+      "learning_rate": 4.9743600601836974e-06,
+      "loss": 1.4292,
+      "step": 21
+    },
+    {
+      "epoch": 0.08088235294117647,
+      "grad_norm": 5.619725929244596,
+      "learning_rate": 4.969917628906234e-06,
+      "loss": 1.3252,
+      "step": 22
+    },
+    {
+      "epoch": 0.08455882352941177,
+      "grad_norm": 5.933112478551199,
+      "learning_rate": 4.96512277356537e-06,
+      "loss": 1.3098,
+      "step": 23
+    },
+    {
+      "epoch": 0.08823529411764706,
+      "grad_norm": 6.037290583894952,
+      "learning_rate": 4.9599761783225465e-06,
+      "loss": 1.4564,
+      "step": 24
+    },
+    {
+      "epoch": 0.09191176470588236,
+      "grad_norm": 6.101449011726076,
+      "learning_rate": 4.954478577527761e-06,
+      "loss": 1.2477,
+      "step": 25
+    },
+    {
+      "epoch": 0.09558823529411764,
+      "grad_norm": 6.484138459877207,
+      "learning_rate": 4.948630755614792e-06,
+      "loss": 1.4502,
+      "step": 26
+    },
+    {
+      "epoch": 0.09926470588235294,
+      "grad_norm": 5.783558029857624,
+      "learning_rate": 4.94243354698926e-06,
+      "loss": 1.2116,
+      "step": 27
+    },
+    {
+      "epoch": 0.10294117647058823,
+      "grad_norm": 5.38825038783534,
+      "learning_rate": 4.935887835909581e-06,
+      "loss": 1.3661,
+      "step": 28
+    },
+    {
+      "epoch": 0.10661764705882353,
+      "grad_norm": 5.885649965944228,
+      "learning_rate": 4.928994556360787e-06,
+      "loss": 1.1933,
+      "step": 29
+    },
+    {
+      "epoch": 0.11029411764705882,
+      "grad_norm": 4.662442963187209,
+      "learning_rate": 4.921754691921262e-06,
+      "loss": 1.154,
+      "step": 30
+    },
+    {
+      "epoch": 0.11397058823529412,
+      "grad_norm": 5.7998445583250815,
+      "learning_rate": 4.914169275622397e-06,
+      "loss": 1.3788,
+      "step": 31
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 6.404293333479338,
+      "learning_rate": 4.906239389801191e-06,
+      "loss": 1.4439,
+      "step": 32
+    },
+    {
+      "epoch": 0.1213235294117647,
+      "grad_norm": 5.687405588659297,
+      "learning_rate": 4.897966165945815e-06,
+      "loss": 1.2126,
+      "step": 33
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 5.015200228303349,
+      "learning_rate": 4.889350784534168e-06,
+      "loss": 1.1298,
+      "step": 34
+    },
+    {
+      "epoch": 0.12867647058823528,
+      "grad_norm": 5.209163642768258,
+      "learning_rate": 4.880394474865433e-06,
+      "loss": 1.157,
+      "step": 35
+    },
+    {
+      "epoch": 0.1323529411764706,
+      "grad_norm": 4.9936448970151,
+      "learning_rate": 4.871098514884675e-06,
+      "loss": 1.0947,
+      "step": 36
+    },
+    {
+      "epoch": 0.13602941176470587,
+      "grad_norm": 4.170158253675973,
+      "learning_rate": 4.8614642310004975e-06,
+      "loss": 0.9862,
+      "step": 37
+    },
+    {
+      "epoch": 0.13970588235294118,
+      "grad_norm": 5.406869110168244,
+      "learning_rate": 4.851492997895777e-06,
+      "loss": 1.2218,
+      "step": 38
+    },
+    {
+      "epoch": 0.14338235294117646,
+      "grad_norm": 4.526741063568119,
+      "learning_rate": 4.841186238331519e-06,
+      "loss": 1.243,
+      "step": 39
+    },
+    {
+      "epoch": 0.14705882352941177,
+      "grad_norm": 4.475264257311105,
+      "learning_rate": 4.830545422943847e-06,
+      "loss": 1.2204,
+      "step": 40
+    },
+    {
+      "epoch": 0.15073529411764705,
+      "grad_norm": 4.477778928179383,
+      "learning_rate": 4.819572070034162e-06,
+      "loss": 1.1599,
+      "step": 41
+    },
+    {
+      "epoch": 0.15441176470588236,
+      "grad_norm": 4.6686458824918295,
+      "learning_rate": 4.808267745352502e-06,
+      "loss": 1.213,
+      "step": 42
+    },
+    {
+      "epoch": 0.15808823529411764,
+      "grad_norm": 4.2722154674052994,
+      "learning_rate": 4.79663406187413e-06,
+      "loss": 1.2486,
+      "step": 43
+    },
+    {
+      "epoch": 0.16176470588235295,
+      "grad_norm": 5.041692672315728,
+      "learning_rate": 4.7846726795693855e-06,
+      "loss": 1.1654,
+      "step": 44
+    },
+    {
+      "epoch": 0.16544117647058823,
+      "grad_norm": 4.269710843388236,
+      "learning_rate": 4.772385305166828e-06,
+      "loss": 1.0043,
+      "step": 45
+    },
+    {
+      "epoch": 0.16911764705882354,
+      "grad_norm": 4.957284051229344,
+      "learning_rate": 4.759773691909708e-06,
+      "loss": 1.0967,
+      "step": 46
+    },
+    {
+      "epoch": 0.17279411764705882,
+      "grad_norm": 4.30286824277265,
+      "learning_rate": 4.746839639305808e-06,
+      "loss": 1.1505,
+      "step": 47
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 4.536322507910887,
+      "learning_rate": 4.733584992870669e-06,
+      "loss": 1.1087,
+      "step": 48
+    },
+    {
+      "epoch": 0.1801470588235294,
+      "grad_norm": 4.843914118708716,
+      "learning_rate": 4.720011643864268e-06,
+      "loss": 1.0722,
+      "step": 49
+    },
+    {
+      "epoch": 0.18382352941176472,
+      "grad_norm": 5.0401878908719295,
+      "learning_rate": 4.706121529021158e-06,
+      "loss": 1.0855,
+      "step": 50
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 4.2875271276562685,
+      "learning_rate": 4.691916630274117e-06,
+      "loss": 0.904,
+      "step": 51
+    },
+    {
+      "epoch": 0.19117647058823528,
+      "grad_norm": 4.43484905950581,
+      "learning_rate": 4.67739897447136e-06,
+      "loss": 1.0189,
+      "step": 52
+    },
+    {
+      "epoch": 0.1948529411764706,
+      "grad_norm": 4.115485449071408,
+      "learning_rate": 4.662570633087333e-06,
+      "loss": 1.0426,
+      "step": 53
+    },
+    {
+      "epoch": 0.19852941176470587,
+      "grad_norm": 5.390871076757842,
+      "learning_rate": 4.647433721927139e-06,
+      "loss": 1.187,
+      "step": 54
+    },
+    {
+      "epoch": 0.20220588235294118,
+      "grad_norm": 4.710723760112048,
+      "learning_rate": 4.631990400824643e-06,
+      "loss": 1.0494,
+      "step": 55
+    },
+    {
+      "epoch": 0.20588235294117646,
+      "grad_norm": 4.280737533882025,
+      "learning_rate": 4.616242873334292e-06,
+      "loss": 0.9632,
+      "step": 56
+    },
+    {
+      "epoch": 0.20955882352941177,
+      "grad_norm": 4.6566814560792285,
+      "learning_rate": 4.600193386416697e-06,
+      "loss": 1.1009,
+      "step": 57
+    },
+    {
+      "epoch": 0.21323529411764705,
+      "grad_norm": 4.75613874105741,
+      "learning_rate": 4.5838442301180245e-06,
+      "loss": 0.9895,
+      "step": 58
+    },
+    {
+      "epoch": 0.21691176470588236,
+      "grad_norm": 4.3159714082125085,
+      "learning_rate": 4.5671977372432355e-06,
+      "loss": 1.2047,
+      "step": 59
+    },
+    {
+      "epoch": 0.22058823529411764,
+      "grad_norm": 3.9786877663369666,
+      "learning_rate": 4.5502562830232225e-06,
+      "loss": 1.0841,
+      "step": 60
+    },
+    {
+      "epoch": 0.22426470588235295,
+      "grad_norm": 4.841963744722847,
+      "learning_rate": 4.533022284775903e-06,
+      "loss": 1.0715,
+      "step": 61
+    },
+    {
+      "epoch": 0.22794117647058823,
+      "grad_norm": 4.663760743549672,
+      "learning_rate": 4.5154982015612965e-06,
+      "loss": 1.0359,
+      "step": 62
+    },
+    {
+      "epoch": 0.23161764705882354,
+      "grad_norm": 4.947384060770783,
+      "learning_rate": 4.497686533830648e-06,
+      "loss": 1.1775,
+      "step": 63
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 4.3251202478818955,
+      "learning_rate": 4.4795898230696535e-06,
+      "loss": 1.0066,
+      "step": 64
+    },
+    {
+      "epoch": 0.23897058823529413,
+      "grad_norm": 4.074916250387871,
+      "learning_rate": 4.461210651435814e-06,
+      "loss": 1.0025,
+      "step": 65
+    },
+    {
+      "epoch": 0.2426470588235294,
+      "grad_norm": 3.9449493974786836,
+      "learning_rate": 4.4425516413900085e-06,
+      "loss": 0.9229,
+      "step": 66
+    },
+    {
+      "epoch": 0.24632352941176472,
+      "grad_norm": 4.30808914514437,
+      "learning_rate": 4.423615455322293e-06,
+      "loss": 1.0541,
+      "step": 67
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.681088253932463,
+      "learning_rate": 4.404404795172022e-06,
+      "loss": 1.1215,
+      "step": 68
+    },
+    {
+      "epoch": 0.2536764705882353,
+      "grad_norm": 4.619074656689764,
+      "learning_rate": 4.38492240204231e-06,
+      "loss": 1.0808,
+      "step": 69
+    },
+    {
+      "epoch": 0.25735294117647056,
+      "grad_norm": 4.4188825963358385,
+      "learning_rate": 4.36517105580892e-06,
+      "loss": 0.9405,
+      "step": 70
+    },
+    {
+      "epoch": 0.2610294117647059,
+      "grad_norm": 4.6261750820403895,
+      "learning_rate": 4.345153574723611e-06,
+      "loss": 1.0477,
+      "step": 71
+    },
+    {
+      "epoch": 0.2647058823529412,
+      "grad_norm": 3.9535524178718093,
+      "learning_rate": 4.324872815012005e-06,
+      "loss": 1.0689,
+      "step": 72
+    },
+    {
+      "epoch": 0.26838235294117646,
+      "grad_norm": 4.360846091272947,
+      "learning_rate": 4.304331670466052e-06,
+      "loss": 1.0726,
+      "step": 73
+    },
+    {
+      "epoch": 0.27205882352941174,
+      "grad_norm": 4.808135374026008,
+      "learning_rate": 4.283533072031116e-06,
+      "loss": 1.0454,
+      "step": 74
+    },
+    {
+      "epoch": 0.2757352941176471,
+      "grad_norm": 4.257447362584399,
+      "learning_rate": 4.262479987387776e-06,
+      "loss": 0.9997,
+      "step": 75
+    },
+    {
+      "epoch": 0.27941176470588236,
+      "grad_norm": 4.3277780528934535,
+      "learning_rate": 4.241175420528369e-06,
+      "loss": 0.9725,
+      "step": 76
+    },
+    {
+      "epoch": 0.28308823529411764,
+      "grad_norm": 4.922824411573383,
+      "learning_rate": 4.21962241132837e-06,
+      "loss": 1.1319,
+      "step": 77
+    },
+    {
+      "epoch": 0.2867647058823529,
+      "grad_norm": 4.812135261686987,
+      "learning_rate": 4.197824035112637e-06,
+      "loss": 1.1085,
+      "step": 78
+    },
+    {
+      "epoch": 0.29044117647058826,
+      "grad_norm": 3.7417895779935937,
+      "learning_rate": 4.175783402216604e-06,
+      "loss": 1.0361,
+      "step": 79
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 4.450944803392342,
+      "learning_rate": 4.153503657542479e-06,
+      "loss": 1.047,
+      "step": 80
+    },
+    {
+      "epoch": 0.2977941176470588,
+      "grad_norm": 5.038399459490228,
+      "learning_rate": 4.130987980110508e-06,
+      "loss": 1.1045,
+      "step": 81
+    },
+    {
+      "epoch": 0.3014705882352941,
+      "grad_norm": 4.269857891095231,
+      "learning_rate": 4.108239582605374e-06,
+      "loss": 0.9607,
+      "step": 82
+    },
+    {
+      "epoch": 0.30514705882352944,
+      "grad_norm": 4.797559209596684,
+      "learning_rate": 4.085261710917786e-06,
+      "loss": 1.0974,
+      "step": 83
+    },
+    {
+      "epoch": 0.3088235294117647,
+      "grad_norm": 3.722091750351139,
+      "learning_rate": 4.062057643681335e-06,
+      "loss": 0.8383,
+      "step": 84
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 4.019708838214477,
+      "learning_rate": 4.0386306918046815e-06,
+      "loss": 0.8893,
+      "step": 85
+    },
+    {
+      "epoch": 0.3161764705882353,
+      "grad_norm": 4.405947058252603,
+      "learning_rate": 4.014984197999125e-06,
+      "loss": 0.9943,
+      "step": 86
+    },
+    {
+      "epoch": 0.31985294117647056,
+      "grad_norm": 4.113754137275219,
+      "learning_rate": 3.991121536301653e-06,
+      "loss": 0.9008,
+      "step": 87
+    },
+    {
+      "epoch": 0.3235294117647059,
+      "grad_norm": 4.055275478204548,
+      "learning_rate": 3.967046111593505e-06,
+      "loss": 1.0629,
+      "step": 88
+    },
+    {
+      "epoch": 0.3272058823529412,
+      "grad_norm": 4.167491368887854,
+      "learning_rate": 3.942761359114345e-06,
+      "loss": 0.8909,
+      "step": 89
+    },
+    {
+      "epoch": 0.33088235294117646,
+      "grad_norm": 3.43259840470699,
+      "learning_rate": 3.918270743972097e-06,
+      "loss": 0.847,
+      "step": 90
+    },
+    {
+      "epoch": 0.33455882352941174,
+      "grad_norm": 5.0261712335509445,
+      "learning_rate": 3.89357776064852e-06,
+      "loss": 0.9767,
+      "step": 91
+    },
+    {
+      "epoch": 0.3382352941176471,
+      "grad_norm": 4.171570657172569,
+      "learning_rate": 3.868685932500596e-06,
+      "loss": 0.9365,
+      "step": 92
+    },
+    {
+      "epoch": 0.34191176470588236,
+      "grad_norm": 4.045384620215937,
+      "learning_rate": 3.84359881125779e-06,
+      "loss": 1.0032,
+      "step": 93
+    },
+    {
+      "epoch": 0.34558823529411764,
+      "grad_norm": 4.0446542340427305,
+      "learning_rate": 3.8183199765152704e-06,
+      "loss": 0.9075,
+      "step": 94
+    },
+    {
+      "epoch": 0.3492647058823529,
+      "grad_norm": 3.907654568146725,
+      "learning_rate": 3.792853035223144e-06,
+      "loss": 0.9802,
+      "step": 95
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 4.037290543250798,
+      "learning_rate": 3.7672016211717977e-06,
+      "loss": 0.9857,
+      "step": 96
+    },
+    {
+      "epoch": 0.35661764705882354,
+      "grad_norm": 4.503123770042846,
+      "learning_rate": 3.7413693944734e-06,
+      "loss": 0.9389,
+      "step": 97
+    },
+    {
+      "epoch": 0.3602941176470588,
+      "grad_norm": 4.156536369995594,
+      "learning_rate": 3.7153600410396558e-06,
+      "loss": 0.8534,
+      "step": 98
+    },
+    {
+      "epoch": 0.3639705882352941,
+      "grad_norm": 4.088503753271803,
+      "learning_rate": 3.689177272055877e-06,
+      "loss": 0.9921,
+      "step": 99
+    },
+    {
+      "epoch": 0.36764705882352944,
+      "grad_norm": 4.741510380256789,
+      "learning_rate": 3.6628248234514434e-06,
+      "loss": 0.9779,
+      "step": 100
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 272,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1876547043328.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

plm/plm3b/checkpoint-100/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64943662d3c18d13ef3a74e38f15699fea6b5918823e24c4df9f3182de637c1f
+size 7224

plm/plm3b/checkpoint-100/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

plm/plm3b/checkpoint-200/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

plm/plm3b/checkpoint-200/config.json

@@ -0,0 +1,65 @@
+{
+  "architectures": [
+    "Qwen2_5_VLForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_token_id": 151655,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 128000,
+  "max_window_layers": 70,
+  "model_type": "qwen2_5_vl",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": {
+    "mrope_section": [
+      16,
+      24,
+      24
+    ],
+    "rope_type": "default",
+    "type": "default"
+  },
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.50.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "video_token_id": 151656,
+  "vision_config": {
+    "depth": 32,
+    "fullatt_block_indexes": [
+      7,
+      15,
+      23,
+      31
+    ],
+    "hidden_act": "silu",
+    "hidden_size": 1280,
+    "in_channels": 3,
+    "in_chans": 3,
+    "intermediate_size": 3420,
+    "model_type": "qwen2_5_vl",
+    "num_heads": 16,
+    "out_hidden_size": 2048,
+    "patch_size": 14,
+    "spatial_merge_size": 2,
+    "spatial_patch_size": 14,
+    "temporal_patch_size": 2,
+    "tokens_per_second": 2,
+    "torch_dtype": "bfloat16",
+    "window_size": 112
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652,
+  "vision_token_id": 151654,
+  "vocab_size": 151936
+}

plm/plm3b/checkpoint-200/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "attn_implementation": "flash_attention_2",
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 1e-06,
+  "transformers_version": "4.50.0"
+}

plm/plm3b/checkpoint-200/global_step200/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94bc49b56d781717e29c10bae73def84b329810bb87868e5502686fa50e9eb5c
+size 37471776486

plm/plm3b/checkpoint-200/global_step200/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:72c2c02881c18311531c8047cfda930554545aecdee43cd014fd7bc733cc6be7
+size 1264465844

plm/plm3b/checkpoint-200/latest

@@ -0,0 +1 @@
+global_step200

plm/plm3b/checkpoint-200/pytorch_model-00001-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccb4e8553691627011e8f8ab3f1dff33658588407b04b2bbfb696280425f8412
+size 4997892186

plm/plm3b/checkpoint-200/pytorch_model-00002-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af48c170e2b23290f316121344cafdd67bc681f41032420bee173775aa013f2d
+size 2511631122

plm/plm3b/checkpoint-200/pytorch_model.bin.index.json

@@ -0,0 +1,832 @@
+{
+  "metadata": {
+    "total_size": 7509245952
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00001-of-00002.bin",
+    "model.embed_tokens.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.20.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.30.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.norm.weight": "pytorch_model-00002-of-00002.bin",
+    "visual.blocks.0.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.ln_q.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.0.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.0.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.2.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.patch_embed.proj.weight": "pytorch_model-00001-of-00002.bin"
+  }
+}

plm/plm3b/checkpoint-200/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:787db515f7b6743e3b35596e1138c851cf5634dd7a63c24ff9474f6ad1af981c
+size 14244

plm/plm3b/checkpoint-200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb42318c029aa2129fb6b714b365674aeb43e90470cf76e9b31c27e80cd2d7cd
+size 1064

plm/plm3b/checkpoint-200/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

plm/plm3b/checkpoint-200/tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

plm/plm3b/checkpoint-200/trainer_state.json

@@ -0,0 +1,1434 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7352941176470589,
+  "eval_steps": 500,
+  "global_step": 200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003676470588235294,
+      "grad_norm": 13.434047998912797,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 2.3324,
+      "step": 1
+    },
+    {
+      "epoch": 0.007352941176470588,
+      "grad_norm": 12.845819001528637,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 2.3146,
+      "step": 2
+    },
+    {
+      "epoch": 0.011029411764705883,
+      "grad_norm": 13.642182825241747,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 2.3436,
+      "step": 3
+    },
+    {
+      "epoch": 0.014705882352941176,
+      "grad_norm": 13.177595637373734,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 2.1726,
+      "step": 4
+    },
+    {
+      "epoch": 0.01838235294117647,
+      "grad_norm": 12.647046172843465,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 2.2284,
+      "step": 5
+    },
+    {
+      "epoch": 0.022058823529411766,
+      "grad_norm": 12.96683330506239,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.251,
+      "step": 6
+    },
+    {
+      "epoch": 0.025735294117647058,
+      "grad_norm": 12.856658525760958,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 2.0926,
+      "step": 7
+    },
+    {
+      "epoch": 0.029411764705882353,
+      "grad_norm": 10.167587096031813,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 2.074,
+      "step": 8
+    },
+    {
+      "epoch": 0.03308823529411765,
+      "grad_norm": 12.228521156970206,
+      "learning_rate": 5e-06,
+      "loss": 1.986,
+      "step": 9
+    },
+    {
+      "epoch": 0.03676470588235294,
+      "grad_norm": 9.687971167815158,
+      "learning_rate": 4.999821641793195e-06,
+      "loss": 2.0738,
+      "step": 10
+    },
+    {
+      "epoch": 0.04044117647058824,
+      "grad_norm": 9.005845290500895,
+      "learning_rate": 4.999286592622096e-06,
+      "loss": 1.7819,
+      "step": 11
+    },
+    {
+      "epoch": 0.04411764705882353,
+      "grad_norm": 7.142208208166851,
+      "learning_rate": 4.998394928831034e-06,
+      "loss": 1.5259,
+      "step": 12
+    },
+    {
+      "epoch": 0.04779411764705882,
+      "grad_norm": 6.173050515660737,
+      "learning_rate": 4.9971467776484526e-06,
+      "loss": 1.5163,
+      "step": 13
+    },
+    {
+      "epoch": 0.051470588235294115,
+      "grad_norm": 6.494634601590806,
+      "learning_rate": 4.995542317168756e-06,
+      "loss": 1.4624,
+      "step": 14
+    },
+    {
+      "epoch": 0.05514705882352941,
+      "grad_norm": 6.170011124682238,
+      "learning_rate": 4.993581776326901e-06,
+      "loss": 1.6327,
+      "step": 15
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 5.756663516704019,
+      "learning_rate": 4.991265434865726e-06,
+      "loss": 1.4856,
+      "step": 16
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 6.442210901739369,
+      "learning_rate": 4.988593623296038e-06,
+      "loss": 1.5692,
+      "step": 17
+    },
+    {
+      "epoch": 0.0661764705882353,
+      "grad_norm": 7.493967445703748,
+      "learning_rate": 4.985566722849454e-06,
+      "loss": 1.6112,
+      "step": 18
+    },
+    {
+      "epoch": 0.06985294117647059,
+      "grad_norm": 6.885056202242848,
+      "learning_rate": 4.9821851654240025e-06,
+      "loss": 1.5994,
+      "step": 19
+    },
+    {
+      "epoch": 0.07352941176470588,
+      "grad_norm": 6.019009530543658,
+      "learning_rate": 4.9784494335225e-06,
+      "loss": 1.4061,
+      "step": 20
+    },
+    {
+      "epoch": 0.07720588235294118,
+      "grad_norm": 7.03669923794518,
+      "learning_rate": 4.9743600601836974e-06,
+      "loss": 1.4292,
+      "step": 21
+    },
+    {
+      "epoch": 0.08088235294117647,
+      "grad_norm": 5.619725929244596,
+      "learning_rate": 4.969917628906234e-06,
+      "loss": 1.3252,
+      "step": 22
+    },
+    {
+      "epoch": 0.08455882352941177,
+      "grad_norm": 5.933112478551199,
+      "learning_rate": 4.96512277356537e-06,
+      "loss": 1.3098,
+      "step": 23
+    },
+    {
+      "epoch": 0.08823529411764706,
+      "grad_norm": 6.037290583894952,
+      "learning_rate": 4.9599761783225465e-06,
+      "loss": 1.4564,
+      "step": 24
+    },
+    {
+      "epoch": 0.09191176470588236,
+      "grad_norm": 6.101449011726076,
+      "learning_rate": 4.954478577527761e-06,
+      "loss": 1.2477,
+      "step": 25
+    },
+    {
+      "epoch": 0.09558823529411764,
+      "grad_norm": 6.484138459877207,
+      "learning_rate": 4.948630755614792e-06,
+      "loss": 1.4502,
+      "step": 26
+    },
+    {
+      "epoch": 0.09926470588235294,
+      "grad_norm": 5.783558029857624,
+      "learning_rate": 4.94243354698926e-06,
+      "loss": 1.2116,
+      "step": 27
+    },
+    {
+      "epoch": 0.10294117647058823,
+      "grad_norm": 5.38825038783534,
+      "learning_rate": 4.935887835909581e-06,
+      "loss": 1.3661,
+      "step": 28
+    },
+    {
+      "epoch": 0.10661764705882353,
+      "grad_norm": 5.885649965944228,
+      "learning_rate": 4.928994556360787e-06,
+      "loss": 1.1933,
+      "step": 29
+    },
+    {
+      "epoch": 0.11029411764705882,
+      "grad_norm": 4.662442963187209,
+      "learning_rate": 4.921754691921262e-06,
+      "loss": 1.154,
+      "step": 30
+    },
+    {
+      "epoch": 0.11397058823529412,
+      "grad_norm": 5.7998445583250815,
+      "learning_rate": 4.914169275622397e-06,
+      "loss": 1.3788,
+      "step": 31
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 6.404293333479338,
+      "learning_rate": 4.906239389801191e-06,
+      "loss": 1.4439,
+      "step": 32
+    },
+    {
+      "epoch": 0.1213235294117647,
+      "grad_norm": 5.687405588659297,
+      "learning_rate": 4.897966165945815e-06,
+      "loss": 1.2126,
+      "step": 33
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 5.015200228303349,
+      "learning_rate": 4.889350784534168e-06,
+      "loss": 1.1298,
+      "step": 34
+    },
+    {
+      "epoch": 0.12867647058823528,
+      "grad_norm": 5.209163642768258,
+      "learning_rate": 4.880394474865433e-06,
+      "loss": 1.157,
+      "step": 35
+    },
+    {
+      "epoch": 0.1323529411764706,
+      "grad_norm": 4.9936448970151,
+      "learning_rate": 4.871098514884675e-06,
+      "loss": 1.0947,
+      "step": 36
+    },
+    {
+      "epoch": 0.13602941176470587,
+      "grad_norm": 4.170158253675973,
+      "learning_rate": 4.8614642310004975e-06,
+      "loss": 0.9862,
+      "step": 37
+    },
+    {
+      "epoch": 0.13970588235294118,
+      "grad_norm": 5.406869110168244,
+      "learning_rate": 4.851492997895777e-06,
+      "loss": 1.2218,
+      "step": 38
+    },
+    {
+      "epoch": 0.14338235294117646,
+      "grad_norm": 4.526741063568119,
+      "learning_rate": 4.841186238331519e-06,
+      "loss": 1.243,
+      "step": 39
+    },
+    {
+      "epoch": 0.14705882352941177,
+      "grad_norm": 4.475264257311105,
+      "learning_rate": 4.830545422943847e-06,
+      "loss": 1.2204,
+      "step": 40
+    },
+    {
+      "epoch": 0.15073529411764705,
+      "grad_norm": 4.477778928179383,
+      "learning_rate": 4.819572070034162e-06,
+      "loss": 1.1599,
+      "step": 41
+    },
+    {
+      "epoch": 0.15441176470588236,
+      "grad_norm": 4.6686458824918295,
+      "learning_rate": 4.808267745352502e-06,
+      "loss": 1.213,
+      "step": 42
+    },
+    {
+      "epoch": 0.15808823529411764,
+      "grad_norm": 4.2722154674052994,
+      "learning_rate": 4.79663406187413e-06,
+      "loss": 1.2486,
+      "step": 43
+    },
+    {
+      "epoch": 0.16176470588235295,
+      "grad_norm": 5.041692672315728,
+      "learning_rate": 4.7846726795693855e-06,
+      "loss": 1.1654,
+      "step": 44
+    },
+    {
+      "epoch": 0.16544117647058823,
+      "grad_norm": 4.269710843388236,
+      "learning_rate": 4.772385305166828e-06,
+      "loss": 1.0043,
+      "step": 45
+    },
+    {
+      "epoch": 0.16911764705882354,
+      "grad_norm": 4.957284051229344,
+      "learning_rate": 4.759773691909708e-06,
+      "loss": 1.0967,
+      "step": 46
+    },
+    {
+      "epoch": 0.17279411764705882,
+      "grad_norm": 4.30286824277265,
+      "learning_rate": 4.746839639305808e-06,
+      "loss": 1.1505,
+      "step": 47
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 4.536322507910887,
+      "learning_rate": 4.733584992870669e-06,
+      "loss": 1.1087,
+      "step": 48
+    },
+    {
+      "epoch": 0.1801470588235294,
+      "grad_norm": 4.843914118708716,
+      "learning_rate": 4.720011643864268e-06,
+      "loss": 1.0722,
+      "step": 49
+    },
+    {
+      "epoch": 0.18382352941176472,
+      "grad_norm": 5.0401878908719295,
+      "learning_rate": 4.706121529021158e-06,
+      "loss": 1.0855,
+      "step": 50
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 4.2875271276562685,
+      "learning_rate": 4.691916630274117e-06,
+      "loss": 0.904,
+      "step": 51
+    },
+    {
+      "epoch": 0.19117647058823528,
+      "grad_norm": 4.43484905950581,
+      "learning_rate": 4.67739897447136e-06,
+      "loss": 1.0189,
+      "step": 52
+    },
+    {
+      "epoch": 0.1948529411764706,
+      "grad_norm": 4.115485449071408,
+      "learning_rate": 4.662570633087333e-06,
+      "loss": 1.0426,
+      "step": 53
+    },
+    {
+      "epoch": 0.19852941176470587,
+      "grad_norm": 5.390871076757842,
+      "learning_rate": 4.647433721927139e-06,
+      "loss": 1.187,
+      "step": 54
+    },
+    {
+      "epoch": 0.20220588235294118,
+      "grad_norm": 4.710723760112048,
+      "learning_rate": 4.631990400824643e-06,
+      "loss": 1.0494,
+      "step": 55
+    },
+    {
+      "epoch": 0.20588235294117646,
+      "grad_norm": 4.280737533882025,
+      "learning_rate": 4.616242873334292e-06,
+      "loss": 0.9632,
+      "step": 56
+    },
+    {
+      "epoch": 0.20955882352941177,
+      "grad_norm": 4.6566814560792285,
+      "learning_rate": 4.600193386416697e-06,
+      "loss": 1.1009,
+      "step": 57
+    },
+    {
+      "epoch": 0.21323529411764705,
+      "grad_norm": 4.75613874105741,
+      "learning_rate": 4.5838442301180245e-06,
+      "loss": 0.9895,
+      "step": 58
+    },
+    {
+      "epoch": 0.21691176470588236,
+      "grad_norm": 4.3159714082125085,
+      "learning_rate": 4.5671977372432355e-06,
+      "loss": 1.2047,
+      "step": 59
+    },
+    {
+      "epoch": 0.22058823529411764,
+      "grad_norm": 3.9786877663369666,
+      "learning_rate": 4.5502562830232225e-06,
+      "loss": 1.0841,
+      "step": 60
+    },
+    {
+      "epoch": 0.22426470588235295,
+      "grad_norm": 4.841963744722847,
+      "learning_rate": 4.533022284775903e-06,
+      "loss": 1.0715,
+      "step": 61
+    },
+    {
+      "epoch": 0.22794117647058823,
+      "grad_norm": 4.663760743549672,
+      "learning_rate": 4.5154982015612965e-06,
+      "loss": 1.0359,
+      "step": 62
+    },
+    {
+      "epoch": 0.23161764705882354,
+      "grad_norm": 4.947384060770783,
+      "learning_rate": 4.497686533830648e-06,
+      "loss": 1.1775,
+      "step": 63
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 4.3251202478818955,
+      "learning_rate": 4.4795898230696535e-06,
+      "loss": 1.0066,
+      "step": 64
+    },
+    {
+      "epoch": 0.23897058823529413,
+      "grad_norm": 4.074916250387871,
+      "learning_rate": 4.461210651435814e-06,
+      "loss": 1.0025,
+      "step": 65
+    },
+    {
+      "epoch": 0.2426470588235294,
+      "grad_norm": 3.9449493974786836,
+      "learning_rate": 4.4425516413900085e-06,
+      "loss": 0.9229,
+      "step": 66
+    },
+    {
+      "epoch": 0.24632352941176472,
+      "grad_norm": 4.30808914514437,
+      "learning_rate": 4.423615455322293e-06,
+      "loss": 1.0541,
+      "step": 67
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.681088253932463,
+      "learning_rate": 4.404404795172022e-06,
+      "loss": 1.1215,
+      "step": 68
+    },
+    {
+      "epoch": 0.2536764705882353,
+      "grad_norm": 4.619074656689764,
+      "learning_rate": 4.38492240204231e-06,
+      "loss": 1.0808,
+      "step": 69
+    },
+    {
+      "epoch": 0.25735294117647056,
+      "grad_norm": 4.4188825963358385,
+      "learning_rate": 4.36517105580892e-06,
+      "loss": 0.9405,
+      "step": 70
+    },
+    {
+      "epoch": 0.2610294117647059,
+      "grad_norm": 4.6261750820403895,
+      "learning_rate": 4.345153574723611e-06,
+      "loss": 1.0477,
+      "step": 71
+    },
+    {
+      "epoch": 0.2647058823529412,
+      "grad_norm": 3.9535524178718093,
+      "learning_rate": 4.324872815012005e-06,
+      "loss": 1.0689,
+      "step": 72
+    },
+    {
+      "epoch": 0.26838235294117646,
+      "grad_norm": 4.360846091272947,
+      "learning_rate": 4.304331670466052e-06,
+      "loss": 1.0726,
+      "step": 73
+    },
+    {
+      "epoch": 0.27205882352941174,
+      "grad_norm": 4.808135374026008,
+      "learning_rate": 4.283533072031116e-06,
+      "loss": 1.0454,
+      "step": 74
+    },
+    {
+      "epoch": 0.2757352941176471,
+      "grad_norm": 4.257447362584399,
+      "learning_rate": 4.262479987387776e-06,
+      "loss": 0.9997,
+      "step": 75
+    },
+    {
+      "epoch": 0.27941176470588236,
+      "grad_norm": 4.3277780528934535,
+      "learning_rate": 4.241175420528369e-06,
+      "loss": 0.9725,
+      "step": 76
+    },
+    {
+      "epoch": 0.28308823529411764,
+      "grad_norm": 4.922824411573383,
+      "learning_rate": 4.21962241132837e-06,
+      "loss": 1.1319,
+      "step": 77
+    },
+    {
+      "epoch": 0.2867647058823529,
+      "grad_norm": 4.812135261686987,
+      "learning_rate": 4.197824035112637e-06,
+      "loss": 1.1085,
+      "step": 78
+    },
+    {
+      "epoch": 0.29044117647058826,
+      "grad_norm": 3.7417895779935937,
+      "learning_rate": 4.175783402216604e-06,
+      "loss": 1.0361,
+      "step": 79
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 4.450944803392342,
+      "learning_rate": 4.153503657542479e-06,
+      "loss": 1.047,
+      "step": 80
+    },
+    {
+      "epoch": 0.2977941176470588,
+      "grad_norm": 5.038399459490228,
+      "learning_rate": 4.130987980110508e-06,
+      "loss": 1.1045,
+      "step": 81
+    },
+    {
+      "epoch": 0.3014705882352941,
+      "grad_norm": 4.269857891095231,
+      "learning_rate": 4.108239582605374e-06,
+      "loss": 0.9607,
+      "step": 82
+    },
+    {
+      "epoch": 0.30514705882352944,
+      "grad_norm": 4.797559209596684,
+      "learning_rate": 4.085261710917786e-06,
+      "loss": 1.0974,
+      "step": 83
+    },
+    {
+      "epoch": 0.3088235294117647,
+      "grad_norm": 3.722091750351139,
+      "learning_rate": 4.062057643681335e-06,
+      "loss": 0.8383,
+      "step": 84
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 4.019708838214477,
+      "learning_rate": 4.0386306918046815e-06,
+      "loss": 0.8893,
+      "step": 85
+    },
+    {
+      "epoch": 0.3161764705882353,
+      "grad_norm": 4.405947058252603,
+      "learning_rate": 4.014984197999125e-06,
+      "loss": 0.9943,
+      "step": 86
+    },
+    {
+      "epoch": 0.31985294117647056,
+      "grad_norm": 4.113754137275219,
+      "learning_rate": 3.991121536301653e-06,
+      "loss": 0.9008,
+      "step": 87
+    },
+    {
+      "epoch": 0.3235294117647059,
+      "grad_norm": 4.055275478204548,
+      "learning_rate": 3.967046111593505e-06,
+      "loss": 1.0629,
+      "step": 88
+    },
+    {
+      "epoch": 0.3272058823529412,
+      "grad_norm": 4.167491368887854,
+      "learning_rate": 3.942761359114345e-06,
+      "loss": 0.8909,
+      "step": 89
+    },
+    {
+      "epoch": 0.33088235294117646,
+      "grad_norm": 3.43259840470699,
+      "learning_rate": 3.918270743972097e-06,
+      "loss": 0.847,
+      "step": 90
+    },
+    {
+      "epoch": 0.33455882352941174,
+      "grad_norm": 5.0261712335509445,
+      "learning_rate": 3.89357776064852e-06,
+      "loss": 0.9767,
+      "step": 91
+    },
+    {
+      "epoch": 0.3382352941176471,
+      "grad_norm": 4.171570657172569,
+      "learning_rate": 3.868685932500596e-06,
+      "loss": 0.9365,
+      "step": 92
+    },
+    {
+      "epoch": 0.34191176470588236,
+      "grad_norm": 4.045384620215937,
+      "learning_rate": 3.84359881125779e-06,
+      "loss": 1.0032,
+      "step": 93
+    },
+    {
+      "epoch": 0.34558823529411764,
+      "grad_norm": 4.0446542340427305,
+      "learning_rate": 3.8183199765152704e-06,
+      "loss": 0.9075,
+      "step": 94
+    },
+    {
+      "epoch": 0.3492647058823529,
+      "grad_norm": 3.907654568146725,
+      "learning_rate": 3.792853035223144e-06,
+      "loss": 0.9802,
+      "step": 95
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 4.037290543250798,
+      "learning_rate": 3.7672016211717977e-06,
+      "loss": 0.9857,
+      "step": 96
+    },
+    {
+      "epoch": 0.35661764705882354,
+      "grad_norm": 4.503123770042846,
+      "learning_rate": 3.7413693944734e-06,
+      "loss": 0.9389,
+      "step": 97
+    },
+    {
+      "epoch": 0.3602941176470588,
+      "grad_norm": 4.156536369995594,
+      "learning_rate": 3.7153600410396558e-06,
+      "loss": 0.8534,
+      "step": 98
+    },
+    {
+      "epoch": 0.3639705882352941,
+      "grad_norm": 4.088503753271803,
+      "learning_rate": 3.689177272055877e-06,
+      "loss": 0.9921,
+      "step": 99
+    },
+    {
+      "epoch": 0.36764705882352944,
+      "grad_norm": 4.741510380256789,
+      "learning_rate": 3.6628248234514434e-06,
+      "loss": 0.9779,
+      "step": 100
+    },
+    {
+      "epoch": 0.3713235294117647,
+      "grad_norm": 4.0308211842614,
+      "learning_rate": 3.6363064553667378e-06,
+      "loss": 0.937,
+      "step": 101
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 4.419077068052694,
+      "learning_rate": 3.6096259516166226e-06,
+      "loss": 0.9745,
+      "step": 102
+    },
+    {
+      "epoch": 0.3786764705882353,
+      "grad_norm": 4.553326993307682,
+      "learning_rate": 3.5827871191505425e-06,
+      "loss": 1.0108,
+      "step": 103
+    },
+    {
+      "epoch": 0.38235294117647056,
+      "grad_norm": 4.560690801720865,
+      "learning_rate": 3.5557937875093242e-06,
+      "loss": 1.1283,
+      "step": 104
+    },
+    {
+      "epoch": 0.3860294117647059,
+      "grad_norm": 3.860973341668436,
+      "learning_rate": 3.528649808278747e-06,
+      "loss": 0.8757,
+      "step": 105
+    },
+    {
+      "epoch": 0.3897058823529412,
+      "grad_norm": 4.171476514896085,
+      "learning_rate": 3.5013590545399818e-06,
+      "loss": 1.0018,
+      "step": 106
+    },
+    {
+      "epoch": 0.39338235294117646,
+      "grad_norm": 4.110319499028112,
+      "learning_rate": 3.473925420316946e-06,
+      "loss": 0.9357,
+      "step": 107
+    },
+    {
+      "epoch": 0.39705882352941174,
+      "grad_norm": 4.145394253919563,
+      "learning_rate": 3.4463528200206868e-06,
+      "loss": 1.0728,
+      "step": 108
+    },
+    {
+      "epoch": 0.4007352941176471,
+      "grad_norm": 4.704663142140023,
+      "learning_rate": 3.4186451878908393e-06,
+      "loss": 1.0716,
+      "step": 109
+    },
+    {
+      "epoch": 0.40441176470588236,
+      "grad_norm": 4.466894626658271,
+      "learning_rate": 3.390806477434269e-06,
+      "loss": 1.0833,
+      "step": 110
+    },
+    {
+      "epoch": 0.40808823529411764,
+      "grad_norm": 3.6939591974327506,
+      "learning_rate": 3.362840660860958e-06,
+      "loss": 0.9237,
+      "step": 111
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 3.933459091229558,
+      "learning_rate": 3.3347517285172225e-06,
+      "loss": 0.8316,
+      "step": 112
+    },
+    {
+      "epoch": 0.41544117647058826,
+      "grad_norm": 4.256779999883048,
+      "learning_rate": 3.3065436883163453e-06,
+      "loss": 0.9603,
+      "step": 113
+    },
+    {
+      "epoch": 0.41911764705882354,
+      "grad_norm": 4.012257242098335,
+      "learning_rate": 3.2782205651667013e-06,
+      "loss": 1.044,
+      "step": 114
+    },
+    {
+      "epoch": 0.4227941176470588,
+      "grad_norm": 5.182250725086201,
+      "learning_rate": 3.2497864003974554e-06,
+      "loss": 1.1498,
+      "step": 115
+    },
+    {
+      "epoch": 0.4264705882352941,
+      "grad_norm": 4.446483805286756,
+      "learning_rate": 3.221245251181919e-06,
+      "loss": 1.0897,
+      "step": 116
+    },
+    {
+      "epoch": 0.43014705882352944,
+      "grad_norm": 3.6215202027140356,
+      "learning_rate": 3.1926011899586483e-06,
+      "loss": 0.8499,
+      "step": 117
+    },
+    {
+      "epoch": 0.4338235294117647,
+      "grad_norm": 4.493062185874119,
+      "learning_rate": 3.1638583038503596e-06,
+      "loss": 1.0812,
+      "step": 118
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 4.604825266791642,
+      "learning_rate": 3.1350206940807523e-06,
+      "loss": 1.2089,
+      "step": 119
+    },
+    {
+      "epoch": 0.4411764705882353,
+      "grad_norm": 4.612080493491828,
+      "learning_rate": 3.10609247538932e-06,
+      "loss": 0.8243,
+      "step": 120
+    },
+    {
+      "epoch": 0.44485294117647056,
+      "grad_norm": 4.124364305045642,
+      "learning_rate": 3.0770777754442333e-06,
+      "loss": 0.9008,
+      "step": 121
+    },
+    {
+      "epoch": 0.4485294117647059,
+      "grad_norm": 3.7931249153443147,
+      "learning_rate": 3.047980734253372e-06,
+      "loss": 0.7878,
+      "step": 122
+    },
+    {
+      "epoch": 0.4522058823529412,
+      "grad_norm": 4.007415833185828,
+      "learning_rate": 3.018805503573612e-06,
+      "loss": 0.8514,
+      "step": 123
+    },
+    {
+      "epoch": 0.45588235294117646,
+      "grad_norm": 4.393069934357272,
+      "learning_rate": 2.989556246318412e-06,
+      "loss": 1.0155,
+      "step": 124
+    },
+    {
+      "epoch": 0.45955882352941174,
+      "grad_norm": 3.890835212583757,
+      "learning_rate": 2.960237135963834e-06,
+      "loss": 0.86,
+      "step": 125
+    },
+    {
+      "epoch": 0.4632352941176471,
+      "grad_norm": 4.694808585667552,
+      "learning_rate": 2.930852355953034e-06,
+      "loss": 1.0037,
+      "step": 126
+    },
+    {
+      "epoch": 0.46691176470588236,
+      "grad_norm": 3.850165960810338,
+      "learning_rate": 2.90140609909935e-06,
+      "loss": 0.8839,
+      "step": 127
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 4.847365116222235,
+      "learning_rate": 2.8719025669880357e-06,
+      "loss": 1.0313,
+      "step": 128
+    },
+    {
+      "epoch": 0.4742647058823529,
+      "grad_norm": 3.050185308576689,
+      "learning_rate": 2.8423459693767586e-06,
+      "loss": 0.823,
+      "step": 129
+    },
+    {
+      "epoch": 0.47794117647058826,
+      "grad_norm": 4.456915112596525,
+      "learning_rate": 2.8127405235949173e-06,
+      "loss": 1.0285,
+      "step": 130
+    },
+    {
+      "epoch": 0.48161764705882354,
+      "grad_norm": 3.853277409686324,
+      "learning_rate": 2.7830904539418884e-06,
+      "loss": 0.8868,
+      "step": 131
+    },
+    {
+      "epoch": 0.4852941176470588,
+      "grad_norm": 3.958874078366555,
+      "learning_rate": 2.7533999910842766e-06,
+      "loss": 0.9671,
+      "step": 132
+    },
+    {
+      "epoch": 0.4889705882352941,
+      "grad_norm": 4.003924135714432,
+      "learning_rate": 2.723673371452254e-06,
+      "loss": 0.724,
+      "step": 133
+    },
+    {
+      "epoch": 0.49264705882352944,
+      "grad_norm": 4.001289329492298,
+      "learning_rate": 2.693914836635076e-06,
+      "loss": 0.9612,
+      "step": 134
+    },
+    {
+      "epoch": 0.4963235294117647,
+      "grad_norm": 4.286343517370892,
+      "learning_rate": 2.664128632775871e-06,
+      "loss": 1.0523,
+      "step": 135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.396122096526358,
+      "learning_rate": 2.634319009965762e-06,
+      "loss": 0.9306,
+      "step": 136
+    },
+    {
+      "epoch": 0.5036764705882353,
+      "grad_norm": 4.578489041399821,
+      "learning_rate": 2.6044902216374497e-06,
+      "loss": 0.9639,
+      "step": 137
+    },
+    {
+      "epoch": 0.5073529411764706,
+      "grad_norm": 3.6914089239089707,
+      "learning_rate": 2.574646523958288e-06,
+      "loss": 0.9325,
+      "step": 138
+    },
+    {
+      "epoch": 0.5110294117647058,
+      "grad_norm": 4.48345328238986,
+      "learning_rate": 2.5447921752230003e-06,
+      "loss": 1.0055,
+      "step": 139
+    },
+    {
+      "epoch": 0.5147058823529411,
+      "grad_norm": 4.416108713410228,
+      "learning_rate": 2.514931435246071e-06,
+      "loss": 0.9215,
+      "step": 140
+    },
+    {
+      "epoch": 0.5183823529411765,
+      "grad_norm": 4.0682969803431215,
+      "learning_rate": 2.48506856475393e-06,
+      "loss": 0.893,
+      "step": 141
+    },
+    {
+      "epoch": 0.5220588235294118,
+      "grad_norm": 3.919681587165293,
+      "learning_rate": 2.4552078247770005e-06,
+      "loss": 0.9927,
+      "step": 142
+    },
+    {
+      "epoch": 0.5257352941176471,
+      "grad_norm": 4.766199064726553,
+      "learning_rate": 2.425353476041713e-06,
+      "loss": 1.0328,
+      "step": 143
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 4.389838168390373,
+      "learning_rate": 2.395509778362552e-06,
+      "loss": 1.0285,
+      "step": 144
+    },
+    {
+      "epoch": 0.5330882352941176,
+      "grad_norm": 3.7114213812482397,
+      "learning_rate": 2.3656809900342383e-06,
+      "loss": 0.9969,
+      "step": 145
+    },
+    {
+      "epoch": 0.5367647058823529,
+      "grad_norm": 4.034177373067454,
+      "learning_rate": 2.33587136722413e-06,
+      "loss": 0.9814,
+      "step": 146
+    },
+    {
+      "epoch": 0.5404411764705882,
+      "grad_norm": 3.69581166677399,
+      "learning_rate": 2.3060851633649246e-06,
+      "loss": 0.8911,
+      "step": 147
+    },
+    {
+      "epoch": 0.5441176470588235,
+      "grad_norm": 4.512622264461683,
+      "learning_rate": 2.2763266285477476e-06,
+      "loss": 1.0292,
+      "step": 148
+    },
+    {
+      "epoch": 0.5477941176470589,
+      "grad_norm": 3.97660109297394,
+      "learning_rate": 2.246600008915724e-06,
+      "loss": 0.8695,
+      "step": 149
+    },
+    {
+      "epoch": 0.5514705882352942,
+      "grad_norm": 3.8939848813303053,
+      "learning_rate": 2.2169095460581116e-06,
+      "loss": 0.8352,
+      "step": 150
+    },
+    {
+      "epoch": 0.5551470588235294,
+      "grad_norm": 3.9870119822321803,
+      "learning_rate": 2.1872594764050835e-06,
+      "loss": 0.909,
+      "step": 151
+    },
+    {
+      "epoch": 0.5588235294117647,
+      "grad_norm": 4.125790157894242,
+      "learning_rate": 2.1576540306232418e-06,
+      "loss": 0.8979,
+      "step": 152
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 4.137257106796478,
+      "learning_rate": 2.1280974330119647e-06,
+      "loss": 1.0781,
+      "step": 153
+    },
+    {
+      "epoch": 0.5661764705882353,
+      "grad_norm": 3.488143252273275,
+      "learning_rate": 2.0985939009006506e-06,
+      "loss": 0.916,
+      "step": 154
+    },
+    {
+      "epoch": 0.5698529411764706,
+      "grad_norm": 3.7818968673565387,
+      "learning_rate": 2.0691476440469674e-06,
+      "loss": 0.9433,
+      "step": 155
+    },
+    {
+      "epoch": 0.5735294117647058,
+      "grad_norm": 3.88076543732921,
+      "learning_rate": 2.0397628640361674e-06,
+      "loss": 0.8835,
+      "step": 156
+    },
+    {
+      "epoch": 0.5772058823529411,
+      "grad_norm": 4.317149769940415,
+      "learning_rate": 2.0104437536815884e-06,
+      "loss": 0.7755,
+      "step": 157
+    },
+    {
+      "epoch": 0.5808823529411765,
+      "grad_norm": 4.237779752503702,
+      "learning_rate": 1.981194496426389e-06,
+      "loss": 0.9332,
+      "step": 158
+    },
+    {
+      "epoch": 0.5845588235294118,
+      "grad_norm": 3.7814283375503615,
+      "learning_rate": 1.9520192657466286e-06,
+      "loss": 0.806,
+      "step": 159
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 4.891793101817214,
+      "learning_rate": 1.9229222245557675e-06,
+      "loss": 0.9093,
+      "step": 160
+    },
+    {
+      "epoch": 0.5919117647058824,
+      "grad_norm": 4.493149810044521,
+      "learning_rate": 1.8939075246106809e-06,
+      "loss": 0.8753,
+      "step": 161
+    },
+    {
+      "epoch": 0.5955882352941176,
+      "grad_norm": 4.8427195938842384,
+      "learning_rate": 1.8649793059192484e-06,
+      "loss": 0.9603,
+      "step": 162
+    },
+    {
+      "epoch": 0.5992647058823529,
+      "grad_norm": 4.387047200700449,
+      "learning_rate": 1.8361416961496412e-06,
+      "loss": 0.8275,
+      "step": 163
+    },
+    {
+      "epoch": 0.6029411764705882,
+      "grad_norm": 4.117080067194527,
+      "learning_rate": 1.8073988100413515e-06,
+      "loss": 1.0691,
+      "step": 164
+    },
+    {
+      "epoch": 0.6066176470588235,
+      "grad_norm": 3.9510884135269557,
+      "learning_rate": 1.7787547488180816e-06,
+      "loss": 0.8831,
+      "step": 165
+    },
+    {
+      "epoch": 0.6102941176470589,
+      "grad_norm": 4.3072942867010715,
+      "learning_rate": 1.7502135996025454e-06,
+      "loss": 1.0632,
+      "step": 166
+    },
+    {
+      "epoch": 0.6139705882352942,
+      "grad_norm": 4.014725966095907,
+      "learning_rate": 1.7217794348332989e-06,
+      "loss": 0.7967,
+      "step": 167
+    },
+    {
+      "epoch": 0.6176470588235294,
+      "grad_norm": 4.701020558501962,
+      "learning_rate": 1.6934563116836555e-06,
+      "loss": 0.955,
+      "step": 168
+    },
+    {
+      "epoch": 0.6213235294117647,
+      "grad_norm": 3.467975781789869,
+      "learning_rate": 1.6652482714827783e-06,
+      "loss": 0.8298,
+      "step": 169
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 4.259901634287036,
+      "learning_rate": 1.6371593391390427e-06,
+      "loss": 0.7926,
+      "step": 170
+    },
+    {
+      "epoch": 0.6286764705882353,
+      "grad_norm": 3.986106417159785,
+      "learning_rate": 1.6091935225657312e-06,
+      "loss": 0.907,
+      "step": 171
+    },
+    {
+      "epoch": 0.6323529411764706,
+      "grad_norm": 4.26670981875997,
+      "learning_rate": 1.581354812109162e-06,
+      "loss": 0.872,
+      "step": 172
+    },
+    {
+      "epoch": 0.6360294117647058,
+      "grad_norm": 4.50532500469795,
+      "learning_rate": 1.553647179979314e-06,
+      "loss": 0.8573,
+      "step": 173
+    },
+    {
+      "epoch": 0.6397058823529411,
+      "grad_norm": 3.5641257364418912,
+      "learning_rate": 1.5260745796830545e-06,
+      "loss": 0.7879,
+      "step": 174
+    },
+    {
+      "epoch": 0.6433823529411765,
+      "grad_norm": 3.829595085981063,
+      "learning_rate": 1.49864094546002e-06,
+      "loss": 0.7794,
+      "step": 175
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 4.102140796578656,
+      "learning_rate": 1.471350191721254e-06,
+      "loss": 0.9274,
+      "step": 176
+    },
+    {
+      "epoch": 0.6507352941176471,
+      "grad_norm": 4.367342451595896,
+      "learning_rate": 1.4442062124906764e-06,
+      "loss": 0.9858,
+      "step": 177
+    },
+    {
+      "epoch": 0.6544117647058824,
+      "grad_norm": 3.9665732009625634,
+      "learning_rate": 1.4172128808494572e-06,
+      "loss": 1.0062,
+      "step": 178
+    },
+    {
+      "epoch": 0.6580882352941176,
+      "grad_norm": 3.7800930903444447,
+      "learning_rate": 1.390374048383379e-06,
+      "loss": 0.8814,
+      "step": 179
+    },
+    {
+      "epoch": 0.6617647058823529,
+      "grad_norm": 3.678472970754693,
+      "learning_rate": 1.3636935446332628e-06,
+      "loss": 0.7775,
+      "step": 180
+    },
+    {
+      "epoch": 0.6654411764705882,
+      "grad_norm": 4.112504648611141,
+      "learning_rate": 1.3371751765485568e-06,
+      "loss": 0.9587,
+      "step": 181
+    },
+    {
+      "epoch": 0.6691176470588235,
+      "grad_norm": 4.163332667518327,
+      "learning_rate": 1.3108227279441243e-06,
+      "loss": 0.879,
+      "step": 182
+    },
+    {
+      "epoch": 0.6727941176470589,
+      "grad_norm": 4.400822072917087,
+      "learning_rate": 1.2846399589603453e-06,
+      "loss": 0.7276,
+      "step": 183
+    },
+    {
+      "epoch": 0.6764705882352942,
+      "grad_norm": 4.053836328068894,
+      "learning_rate": 1.2586306055266007e-06,
+      "loss": 0.9748,
+      "step": 184
+    },
+    {
+      "epoch": 0.6801470588235294,
+      "grad_norm": 3.1311278376791285,
+      "learning_rate": 1.2327983788282033e-06,
+      "loss": 0.6834,
+      "step": 185
+    },
+    {
+      "epoch": 0.6838235294117647,
+      "grad_norm": 4.093902627154914,
+      "learning_rate": 1.2071469647768564e-06,
+      "loss": 0.8092,
+      "step": 186
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 3.8257305935130987,
+      "learning_rate": 1.1816800234847304e-06,
+      "loss": 0.908,
+      "step": 187
+    },
+    {
+      "epoch": 0.6911764705882353,
+      "grad_norm": 3.4506718543608654,
+      "learning_rate": 1.1564011887422098e-06,
+      "loss": 0.8306,
+      "step": 188
+    },
+    {
+      "epoch": 0.6948529411764706,
+      "grad_norm": 4.300603297700846,
+      "learning_rate": 1.1313140674994053e-06,
+      "loss": 1.0377,
+      "step": 189
+    },
+    {
+      "epoch": 0.6985294117647058,
+      "grad_norm": 4.200147063692658,
+      "learning_rate": 1.106422239351481e-06,
+      "loss": 0.9722,
+      "step": 190
+    },
+    {
+      "epoch": 0.7022058823529411,
+      "grad_norm": 4.595993638176125,
+      "learning_rate": 1.0817292560279038e-06,
+      "loss": 1.0687,
+      "step": 191
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 3.849489179194865,
+      "learning_rate": 1.0572386408856553e-06,
+      "loss": 0.9187,
+      "step": 192
+    },
+    {
+      "epoch": 0.7095588235294118,
+      "grad_norm": 4.407775963916882,
+      "learning_rate": 1.0329538884064948e-06,
+      "loss": 0.9778,
+      "step": 193
+    },
+    {
+      "epoch": 0.7132352941176471,
+      "grad_norm": 3.6432430902681574,
+      "learning_rate": 1.0088784636983472e-06,
+      "loss": 0.9334,
+      "step": 194
+    },
+    {
+      "epoch": 0.7169117647058824,
+      "grad_norm": 4.452522834485264,
+      "learning_rate": 9.850158020008757e-07,
+      "loss": 0.8688,
+      "step": 195
+    },
+    {
+      "epoch": 0.7205882352941176,
+      "grad_norm": 3.965775509729144,
+      "learning_rate": 9.613693081953195e-07,
+      "loss": 0.9423,
+      "step": 196
+    },
+    {
+      "epoch": 0.7242647058823529,
+      "grad_norm": 4.309078365985055,
+      "learning_rate": 9.379423563186652e-07,
+      "loss": 0.9924,
+      "step": 197
+    },
+    {
+      "epoch": 0.7279411764705882,
+      "grad_norm": 4.0361129691155915,
+      "learning_rate": 9.147382890822143e-07,
+      "loss": 0.9331,
+      "step": 198
+    },
+    {
+      "epoch": 0.7316176470588235,
+      "grad_norm": 3.671894446314836,
+      "learning_rate": 8.917604173946268e-07,
+      "loss": 0.9175,
+      "step": 199
+    },
+    {
+      "epoch": 0.7352941176470589,
+      "grad_norm": 3.6656991490850643,
+      "learning_rate": 8.690120198894919e-07,
+      "loss": 0.7873,
+      "step": 200
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 272,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3765506621440.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

plm/plm3b/checkpoint-200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64943662d3c18d13ef3a74e38f15699fea6b5918823e24c4df9f3182de637c1f
+size 7224

plm/plm3b/checkpoint-200/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

plm/plm3b/checkpoint-272/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

plm/plm3b/checkpoint-272/config.json

@@ -0,0 +1,65 @@
+{
+  "architectures": [
+    "Qwen2_5_VLForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_token_id": 151655,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 128000,
+  "max_window_layers": 70,
+  "model_type": "qwen2_5_vl",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": {
+    "mrope_section": [
+      16,
+      24,
+      24
+    ],
+    "rope_type": "default",
+    "type": "default"
+  },
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.50.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "video_token_id": 151656,
+  "vision_config": {
+    "depth": 32,
+    "fullatt_block_indexes": [
+      7,
+      15,
+      23,
+      31
+    ],
+    "hidden_act": "silu",
+    "hidden_size": 1280,
+    "in_channels": 3,
+    "in_chans": 3,
+    "intermediate_size": 3420,
+    "model_type": "qwen2_5_vl",
+    "num_heads": 16,
+    "out_hidden_size": 2048,
+    "patch_size": 14,
+    "spatial_merge_size": 2,
+    "spatial_patch_size": 14,
+    "temporal_patch_size": 2,
+    "tokens_per_second": 2,
+    "torch_dtype": "bfloat16",
+    "window_size": 112
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652,
+  "vision_token_id": 151654,
+  "vocab_size": 151936
+}

plm/plm3b/checkpoint-272/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "attn_implementation": "flash_attention_2",
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 1e-06,
+  "transformers_version": "4.50.0"
+}

plm/plm3b/checkpoint-272/global_step272/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b18b52522d197e77df7a5c74ef85a94938a8351421d8b4dbdd567005c20826ff
+size 37471776486

plm/plm3b/checkpoint-272/global_step272/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b86c63931f27d9e9acd7160a7da71c869963d4202290243d7354a9411ed0e490
+size 1264465844

plm/plm3b/checkpoint-272/latest

@@ -0,0 +1 @@
+global_step272

plm/plm3b/checkpoint-272/pytorch_model-00001-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2c10f4c9aee05bb455ac9344dddc4941bf4cd0984c2445b0c28638016c1265c
+size 4997892186

plm/plm3b/checkpoint-272/pytorch_model-00002-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab59ac6968a63cfd7205ed61143a917afc9dfd79568a9417a1fe696a7e5711a7
+size 2511631122

plm/plm3b/checkpoint-272/pytorch_model.bin.index.json

@@ -0,0 +1,832 @@
+{
+  "metadata": {
+    "total_size": 7509245952
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00001-of-00002.bin",
+    "model.embed_tokens.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.20.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.30.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.32.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.33.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.34.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.k_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.q_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.v_proj.bias": "pytorch_model-00002-of-00002.bin",
+    "model.layers.35.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.k_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.q_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.v_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.norm.weight": "pytorch_model-00002-of-00002.bin",
+    "visual.blocks.0.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.0.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.1.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.10.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.11.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.12.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.13.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.14.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.15.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.16.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.17.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.18.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.19.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.2.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.20.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.21.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.22.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.23.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.24.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.25.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.26.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.27.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.28.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.29.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.3.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.30.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.31.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.4.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.5.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.6.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.7.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.8.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.qkv.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.attn.qkv.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.down_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.gate_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.up_proj.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.norm1.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.blocks.9.norm2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.ln_q.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.0.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.0.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.2.bias": "pytorch_model-00001-of-00002.bin",
+    "visual.merger.mlp.2.weight": "pytorch_model-00001-of-00002.bin",
+    "visual.patch_embed.proj.weight": "pytorch_model-00001-of-00002.bin"
+  }
+}

plm/plm3b/checkpoint-272/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:787db515f7b6743e3b35596e1138c851cf5634dd7a63c24ff9474f6ad1af981c
+size 14244

plm/plm3b/checkpoint-272/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:322f645560e33b1b1d7e0b7cecda2455f5307484068f6d25ca9f6501684aaa70
+size 1064