Upload MptForCausalLM

config.json

@@ -0,0 +1,118 @@
+{
+  "_commit_hash": "72e5f594ce36f9cabfa2a9fd8f58b491eb467ee7",
+  "_name_or_path": "mosaicml/mpt-7b",
+  "architectures": [
+    "MptForCausalLM"
+  ],
+  "attn_config": {
+    "_name_or_path": "",
+    "add_cross_attention": false,
+    "alibi": true,
+    "alibi_bias_max": 8,
+    "architectures": null,
+    "attn_impl": "torch",
+    "attn_pdrop": 0,
+    "attn_type": "multihead_attention",
+    "attn_uses_sequence_id": false,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "clip_qkv": null,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "min_length": 0,
+    "model_type": "",
+    "no_repeat_ngram_size": 0,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": null,
+    "prefix": null,
+    "prefix_lm": false,
+    "problem_type": null,
+    "pruned_heads": {},
+    "qk_ln": false,
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "softmax_scale": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": null,
+    "torchscript": false,
+    "transformers_version": "4.32.0.dev0",
+    "typical_p": 1.0,
+    "use_bfloat16": false
+  },
+  "auto_map": {
+    "AutoConfig": "mosaicml/mpt-7b--configuration_mpt.MPTConfig",
+    "AutoModelForCausalLM": "mosaicml/mpt-7b--modeling_mpt.MPTForCausalLM"
+  },
+  "d_model": 4096,
+  "emb_pdrop": 0,
+  "embedding_fraction": 1.0,
+  "expansion_ratio": 4,
+  "init_config": {
+    "emb_init_std": null,
+    "emb_init_uniform_lim": null,
+    "fan_mode": "fan_in",
+    "init_div_is_residual": true,
+    "init_gain": 0,
+    "init_nonlinearity": "relu",
+    "init_std": 0.02,
+    "name": "kaiming_normal_",
+    "verbose": 0
+  },
+  "init_device": "cpu",
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "learned_pos_emb": true,
+  "logit_scale": null,
+  "max_seq_len": 2048,
+  "model_type": "mpt",
+  "n_heads": 32,
+  "n_layers": 32,
+  "no_bias": true,
+  "norm_type": "low_precision_layernorm",
+  "resid_pdrop": 0,
+  "tokenizer_name": "EleutherAI/gpt-neox-20b",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.28.1",
+  "use_cache": false,
+  "verbose": 0,
+  "vocab_size": 50432
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "eos_token_id": 0,
+  "transformers_version": "4.32.0.dev0",
+  "use_cache": false
+}

pytorch_model-00001-of-00007.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8880f26aaa46cc2e06ff8b3377f25c5841b49298d4e88804af40fc96e87d3cd4
+size 1889607935

pytorch_model-00002-of-00007.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34ffe56e54aa3c9e0574fbe90840041a7860b3edbd41d04f73794456174c57b1
+size 1879140121

pytorch_model-00003-of-00007.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0118c986b7fa1dc7f862fabe0c8404cbf226ef496b31e49d951d21acb678270a
+size 1979794887

pytorch_model-00004-of-00007.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4688b0c69dd2714aef19bbbd9ac85f01d10535f64dc4275b68067a9472db3506
+size 1912694553

pytorch_model-00005-of-00007.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd0147e4ba0c8f02dd09b578d3f1c95018c99b20e117be1f59be1f53ba982f0c
+size 1879131591

pytorch_model-00006-of-00007.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e260ae12dac8d6d2952eead92e24df96297a63bb64175e048391d223753cc495
+size 1879140121

pytorch_model-00007-of-00007.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:131ca3760ef5e162998b00b6dc666dabf84d42276f184f3a5a1e4d1bf394e753
+size 1879131253

pytorch_model.bin.index.json

@@ -0,0 +1,202 @@
+{
+  "metadata": {
+    "total_size": 13298573312
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.0.attn.Wqkv.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.0.attn.out_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.0.ffn.down_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.0.ffn.up_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.0.norm_1.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.0.norm_2.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.1.attn.Wqkv.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.1.attn.out_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.1.ffn.down_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.1.ffn.up_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.1.norm_1.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.1.norm_2.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.10.attn.Wqkv.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.10.attn.out_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.10.ffn.down_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.10.ffn.up_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.10.norm_1.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.10.norm_2.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.11.attn.Wqkv.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.11.attn.out_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.11.ffn.down_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.11.ffn.up_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.11.norm_1.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.11.norm_2.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.12.attn.Wqkv.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.12.attn.out_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.12.ffn.down_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.12.ffn.up_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.12.norm_1.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.12.norm_2.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.13.attn.Wqkv.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.13.attn.out_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.13.ffn.down_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.13.ffn.up_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.13.norm_1.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.13.norm_2.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.14.attn.Wqkv.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.14.attn.out_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.14.ffn.down_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.14.ffn.up_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.14.norm_1.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.14.norm_2.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.15.attn.Wqkv.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.15.attn.out_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.15.ffn.down_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.15.ffn.up_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.15.norm_1.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.15.norm_2.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.16.attn.Wqkv.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.16.attn.out_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.16.ffn.down_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.16.ffn.up_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.16.norm_1.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.16.norm_2.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.17.attn.Wqkv.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.17.attn.out_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.17.ffn.down_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.17.ffn.up_proj.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.17.norm_1.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.17.norm_2.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.18.attn.Wqkv.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.18.attn.out_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.18.ffn.down_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.18.ffn.up_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.18.norm_1.weight": "pytorch_model-00004-of-00007.bin",
+    "transformer.blocks.18.norm_2.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.19.attn.Wqkv.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.19.attn.out_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.19.ffn.down_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.19.ffn.up_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.19.norm_1.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.19.norm_2.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.2.attn.Wqkv.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.2.attn.out_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.2.ffn.down_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.2.ffn.up_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.2.norm_1.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.2.norm_2.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.20.attn.Wqkv.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.20.attn.out_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.20.ffn.down_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.20.ffn.up_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.20.norm_1.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.20.norm_2.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.21.attn.Wqkv.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.21.attn.out_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.21.ffn.down_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.21.ffn.up_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.21.norm_1.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.21.norm_2.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.22.attn.Wqkv.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.22.attn.out_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.22.ffn.down_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.22.ffn.up_proj.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.22.norm_1.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.22.norm_2.weight": "pytorch_model-00005-of-00007.bin",
+    "transformer.blocks.23.attn.Wqkv.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.23.attn.out_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.23.ffn.down_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.23.ffn.up_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.23.norm_1.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.23.norm_2.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.24.attn.Wqkv.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.24.attn.out_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.24.ffn.down_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.24.ffn.up_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.24.norm_1.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.24.norm_2.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.25.attn.Wqkv.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.25.attn.out_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.25.ffn.down_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.25.ffn.up_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.25.norm_1.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.25.norm_2.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.26.attn.Wqkv.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.26.attn.out_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.26.ffn.down_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.26.ffn.up_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.26.norm_1.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.26.norm_2.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.27.attn.Wqkv.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.27.attn.out_proj.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.27.ffn.down_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.27.ffn.up_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.27.norm_1.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.27.norm_2.weight": "pytorch_model-00006-of-00007.bin",
+    "transformer.blocks.28.attn.Wqkv.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.28.attn.out_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.28.ffn.down_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.28.ffn.up_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.28.norm_1.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.28.norm_2.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.29.attn.Wqkv.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.29.attn.out_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.29.ffn.down_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.29.ffn.up_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.29.norm_1.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.29.norm_2.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.3.attn.Wqkv.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.3.attn.out_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.3.ffn.down_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.3.ffn.up_proj.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.3.norm_1.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.3.norm_2.weight": "pytorch_model-00001-of-00007.bin",
+    "transformer.blocks.30.attn.Wqkv.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.30.attn.out_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.30.ffn.down_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.30.ffn.up_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.30.norm_1.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.30.norm_2.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.31.attn.Wqkv.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.31.attn.out_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.31.ffn.down_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.31.ffn.up_proj.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.31.norm_1.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.31.norm_2.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.blocks.4.attn.Wqkv.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.4.attn.out_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.4.ffn.down_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.4.ffn.up_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.4.norm_1.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.4.norm_2.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.5.attn.Wqkv.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.5.attn.out_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.5.ffn.down_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.5.ffn.up_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.5.norm_1.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.5.norm_2.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.6.attn.Wqkv.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.6.attn.out_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.6.ffn.down_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.6.ffn.up_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.6.norm_1.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.6.norm_2.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.7.attn.Wqkv.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.7.attn.out_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.7.ffn.down_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.7.ffn.up_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.7.norm_1.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.7.norm_2.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.8.attn.Wqkv.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.8.attn.out_proj.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.8.ffn.down_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.8.ffn.up_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.8.norm_1.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.8.norm_2.weight": "pytorch_model-00002-of-00007.bin",
+    "transformer.blocks.9.attn.Wqkv.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.9.attn.out_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.9.ffn.down_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.9.ffn.up_proj.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.9.norm_1.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.blocks.9.norm_2.weight": "pytorch_model-00003-of-00007.bin",
+    "transformer.norm_f.weight": "pytorch_model-00007-of-00007.bin",
+    "transformer.wte.weight": "pytorch_model-00001-of-00007.bin"
+  }
+}