Upload MPTForCausalLM

.gitattributes

@@ -0,0 +1,7 @@
+model-00001-of-00007.safetensors filter=lfs diff=lfs merge=lfs -text
+model-00002-of-00007.safetensors filter=lfs diff=lfs merge=lfs -text
+model-00003-of-00007.safetensors filter=lfs diff=lfs merge=lfs -text
+model-00004-of-00007.safetensors filter=lfs diff=lfs merge=lfs -text
+model-00005-of-00007.safetensors filter=lfs diff=lfs merge=lfs -text
+model-00006-of-00007.safetensors filter=lfs diff=lfs merge=lfs -text
+model-00007-of-00007.safetensors filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,8 +1,8 @@
 ---
-license: mit
 language:
 - th
 - en
+license: mit
 base_model: aisingapore/sea-lion-7b-instruct
 datasets:
 - AIAT/Optimizer-datasetfinal

config.json

@@ -0,0 +1,74 @@
+{
+  "_name_or_path": "aisingapore/sea-lion-7b-instruct",
+  "architectures": [
+    "MPTForCausalLM"
+  ],
+  "attn_config": {
+    "alibi": false,
+    "alibi_bias_max": 8,
+    "attn_impl": "torch",
+    "attn_pdrop": 0.0,
+    "attn_type": "multihead_attention",
+    "attn_uses_sequence_id": false,
+    "clip_qkv": null,
+    "prefix_lm": false,
+    "qk_gn": false,
+    "qk_ln": true,
+    "rope": false,
+    "rope_dail_config": {
+      "pos_idx_in_fp32": true,
+      "type": "original",
+      "xpos_scale_base": 512
+    },
+    "rope_hf_config": {
+      "factor": 1.0,
+      "type": "no_scaling"
+    },
+    "rope_impl": "dail",
+    "rope_theta": 10000,
+    "sliding_window_size": -1,
+    "softmax_scale": null
+  },
+  "auto_map": {
+    "AutoConfig": "aisingapore/sea-lion-7b-instruct--configuration_mpt.MPTConfig",
+    "AutoModelForCausalLM": "aisingapore/sea-lion-7b-instruct--modeling_mpt.MPTForCausalLM"
+  },
+  "d_model": 4096,
+  "emb_pdrop": 0.0,
+  "embedding_fraction": 0.1,
+  "expansion_ratio": 4,
+  "fc_type": "torch",
+  "ffn_config": {
+    "fc_type": "torch",
+    "ffn_type": "mptmlp"
+  },
+  "init_config": {
+    "emb_init_std": null,
+    "emb_init_uniform_lim": null,
+    "fan_mode": "fan_in",
+    "init_div_is_residual": true,
+    "init_gain": 0.0,
+    "init_nonlinearity": "relu",
+    "init_std": null,
+    "name": "kaiming_normal_",
+    "verbose": 0
+  },
+  "init_config_defaults": {
+    "init_std": 0.02
+  },
+  "init_device": "cpu",
+  "learned_pos_emb": true,
+  "logit_scale": "inv_sqrt_d_model",
+  "max_seq_len": 2048,
+  "model_type": "mpt",
+  "n_heads": 32,
+  "n_layers": 32,
+  "no_bias": false,
+  "norm_type": "low_precision_layernorm",
+  "resid_pdrop": 0.0,
+  "torch_dtype": "float32",
+  "transformers_version": "4.40.1",
+  "use_cache": false,
+  "use_pad_tok_in_ffn": true,
+  "vocab_size": 256000
+}

generation_config.json

@@ -0,0 +1,4 @@
+{
+  "_from_model_config": true,
+  "transformers_version": "4.40.1"
+}

model-00001-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ddef36625a8232030cf0bc932dc377fc8bb3a669c60c476777730d59f28873a
+size 4764993232

model-00002-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:341515c1fca75f5b50185cb594d5c83d91f99e242df15618b511500ad84412ee
+size 4833520040

model-00003-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21e9e53cc476a201953a5776afe4fde827794fd7d803623880d8118ef267b9eb
+size 4833520080

model-00004-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94598a3c5e6134d70f99f7a4c1b29367a9dcd3480709e0ae288c6ade77b538cb
+size 4833520136

model-00005-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:971facdbf88127bed1cf74ee0165bc0ba8aa1e03b80298b1c28a78b984c3fcd9
+size 4833520136

model-00006-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5100f46877ddd8fe1dae8249b871beb7994afb47e0e4fbc2a543f434e2c6cbe
+size 4833520136

model-00007-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79803387e796bab81aaaea595137c3af29a352fc89ee3d8676a6436d42afa5d2
+size 1074071728

model.safetensors.index.json

@@ -0,0 +1,523 @@
+{
+  "metadata": {
+    "total_size": 30006607872
+  },
+  "weight_map": {
+    "transformer.blocks.0.attn.Wqkv.bias": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.attn.Wqkv.weight": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.attn.k_ln.bias": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.attn.k_ln.weight": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.attn.out_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.attn.out_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.attn.q_ln.bias": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.attn.q_ln.weight": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.ffn.down_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.0.ffn.down_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.0.ffn.up_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.ffn.up_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.norm_1.bias": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.norm_1.weight": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.norm_2.bias": "model-00001-of-00007.safetensors",
+    "transformer.blocks.0.norm_2.weight": "model-00001-of-00007.safetensors",
+    "transformer.blocks.1.attn.Wqkv.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.attn.Wqkv.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.attn.k_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.attn.k_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.attn.out_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.attn.out_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.attn.q_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.attn.q_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.ffn.down_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.ffn.down_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.ffn.up_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.ffn.up_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.norm_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.norm_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.norm_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.1.norm_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.10.attn.Wqkv.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.attn.Wqkv.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.attn.k_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.attn.k_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.attn.out_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.attn.out_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.attn.q_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.attn.q_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.ffn.down_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.ffn.down_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.ffn.up_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.ffn.up_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.norm_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.norm_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.norm_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.10.norm_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.Wqkv.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.Wqkv.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.k_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.k_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.out_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.out_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.q_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.attn.q_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.ffn.down_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.ffn.down_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.ffn.up_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.ffn.up_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.norm_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.norm_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.norm_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.11.norm_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.Wqkv.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.Wqkv.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.k_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.k_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.out_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.out_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.q_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.attn.q_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.ffn.down_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.12.ffn.down_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.12.ffn.up_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.ffn.up_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.norm_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.norm_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.norm_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.12.norm_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.13.attn.Wqkv.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.attn.Wqkv.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.attn.k_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.attn.k_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.attn.out_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.attn.out_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.attn.q_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.attn.q_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.ffn.down_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.ffn.down_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.ffn.up_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.ffn.up_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.norm_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.norm_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.norm_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.13.norm_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.Wqkv.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.Wqkv.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.k_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.k_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.out_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.out_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.q_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.attn.q_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.ffn.down_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.ffn.down_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.ffn.up_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.ffn.up_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.norm_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.norm_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.norm_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.14.norm_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.Wqkv.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.Wqkv.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.k_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.k_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.out_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.out_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.q_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.attn.q_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.ffn.down_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.ffn.down_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.ffn.up_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.ffn.up_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.norm_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.norm_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.norm_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.15.norm_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.Wqkv.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.Wqkv.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.k_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.k_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.out_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.out_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.q_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.attn.q_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.ffn.down_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.ffn.down_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.ffn.up_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.ffn.up_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.norm_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.norm_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.norm_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.16.norm_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.Wqkv.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.Wqkv.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.k_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.k_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.out_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.out_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.q_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.attn.q_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.ffn.down_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.ffn.down_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.ffn.up_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.ffn.up_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.norm_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.norm_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.norm_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.17.norm_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.Wqkv.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.Wqkv.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.k_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.k_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.out_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.out_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.q_ln.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.attn.q_ln.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.ffn.down_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.18.ffn.down_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.18.ffn.up_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.ffn.up_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.norm_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.norm_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.norm_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.blocks.18.norm_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.blocks.19.attn.Wqkv.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.attn.Wqkv.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.attn.k_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.attn.k_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.attn.out_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.attn.out_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.attn.q_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.attn.q_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.ffn.down_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.ffn.down_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.ffn.up_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.ffn.up_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.norm_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.norm_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.norm_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.19.norm_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.2.attn.Wqkv.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.attn.Wqkv.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.attn.k_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.attn.k_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.attn.out_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.attn.out_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.attn.q_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.attn.q_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.ffn.down_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.ffn.down_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.ffn.up_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.ffn.up_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.norm_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.norm_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.norm_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.2.norm_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.20.attn.Wqkv.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.attn.Wqkv.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.attn.k_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.attn.k_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.attn.out_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.attn.out_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.attn.q_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.attn.q_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.ffn.down_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.ffn.down_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.ffn.up_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.ffn.up_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.norm_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.norm_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.norm_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.20.norm_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.Wqkv.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.Wqkv.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.k_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.k_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.out_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.out_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.q_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.attn.q_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.ffn.down_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.ffn.down_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.ffn.up_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.ffn.up_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.norm_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.norm_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.norm_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.21.norm_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.Wqkv.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.Wqkv.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.k_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.k_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.out_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.out_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.q_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.attn.q_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.ffn.down_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.ffn.down_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.ffn.up_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.ffn.up_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.norm_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.norm_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.norm_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.22.norm_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.Wqkv.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.Wqkv.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.k_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.k_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.out_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.out_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.q_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.attn.q_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.ffn.down_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.ffn.down_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.ffn.up_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.ffn.up_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.norm_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.norm_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.norm_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.23.norm_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.Wqkv.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.Wqkv.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.k_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.k_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.out_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.out_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.q_ln.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.attn.q_ln.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.ffn.down_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.24.ffn.down_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.24.ffn.up_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.ffn.up_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.norm_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.norm_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.norm_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.blocks.24.norm_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.blocks.25.attn.Wqkv.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.attn.Wqkv.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.attn.k_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.attn.k_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.attn.out_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.attn.out_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.attn.q_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.attn.q_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.ffn.down_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.ffn.down_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.ffn.up_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.ffn.up_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.norm_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.norm_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.norm_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.25.norm_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.Wqkv.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.Wqkv.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.k_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.k_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.out_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.out_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.q_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.attn.q_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.ffn.down_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.ffn.down_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.ffn.up_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.ffn.up_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.norm_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.norm_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.norm_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.26.norm_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.Wqkv.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.Wqkv.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.k_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.k_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.out_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.out_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.q_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.attn.q_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.ffn.down_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.ffn.down_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.ffn.up_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.ffn.up_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.norm_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.norm_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.norm_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.27.norm_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.Wqkv.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.Wqkv.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.k_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.k_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.out_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.out_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.q_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.attn.q_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.ffn.down_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.ffn.down_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.ffn.up_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.ffn.up_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.norm_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.norm_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.norm_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.28.norm_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.Wqkv.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.Wqkv.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.k_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.k_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.out_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.out_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.q_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.attn.q_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.ffn.down_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.ffn.down_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.ffn.up_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.ffn.up_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.norm_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.norm_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.norm_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.29.norm_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.3.attn.Wqkv.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.attn.Wqkv.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.attn.k_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.attn.k_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.attn.out_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.attn.out_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.attn.q_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.attn.q_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.ffn.down_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.ffn.down_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.ffn.up_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.ffn.up_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.norm_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.norm_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.norm_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.3.norm_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.30.attn.Wqkv.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.attn.Wqkv.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.attn.k_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.attn.k_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.attn.out_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.attn.out_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.attn.q_ln.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.attn.q_ln.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.ffn.down_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.30.ffn.down_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.30.ffn.up_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.ffn.up_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.norm_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.norm_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.norm_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.blocks.30.norm_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.blocks.31.attn.Wqkv.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.attn.Wqkv.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.attn.k_ln.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.attn.k_ln.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.attn.out_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.attn.out_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.attn.q_ln.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.attn.q_ln.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.ffn.down_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.ffn.down_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.ffn.up_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.ffn.up_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.norm_1.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.norm_1.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.norm_2.bias": "model-00007-of-00007.safetensors",
+    "transformer.blocks.31.norm_2.weight": "model-00007-of-00007.safetensors",
+    "transformer.blocks.4.attn.Wqkv.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.attn.Wqkv.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.attn.k_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.attn.k_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.attn.out_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.attn.out_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.attn.q_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.attn.q_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.ffn.down_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.ffn.down_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.ffn.up_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.ffn.up_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.norm_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.norm_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.norm_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.4.norm_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.Wqkv.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.Wqkv.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.k_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.k_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.out_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.out_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.q_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.attn.q_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.ffn.down_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.ffn.down_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.ffn.up_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.ffn.up_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.norm_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.norm_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.norm_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.5.norm_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.Wqkv.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.Wqkv.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.k_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.k_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.out_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.out_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.q_ln.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.attn.q_ln.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.ffn.down_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.6.ffn.down_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.6.ffn.up_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.ffn.up_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.norm_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.norm_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.norm_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.blocks.6.norm_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.blocks.7.attn.Wqkv.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.attn.Wqkv.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.attn.k_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.attn.k_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.attn.out_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.attn.out_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.attn.q_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.attn.q_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.ffn.down_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.ffn.down_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.ffn.up_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.ffn.up_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.norm_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.norm_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.norm_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.7.norm_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.Wqkv.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.Wqkv.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.k_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.k_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.out_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.out_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.q_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.attn.q_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.ffn.down_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.ffn.down_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.ffn.up_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.ffn.up_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.norm_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.norm_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.norm_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.8.norm_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.Wqkv.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.Wqkv.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.k_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.k_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.out_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.out_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.q_ln.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.attn.q_ln.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.ffn.down_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.ffn.down_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.ffn.up_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.ffn.up_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.norm_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.norm_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.norm_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.blocks.9.norm_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.norm_f.bias": "model-00007-of-00007.safetensors",
+    "transformer.norm_f.weight": "model-00007-of-00007.safetensors",
+    "transformer.wpe.weight": "model-00001-of-00007.safetensors",
+    "transformer.wte.weight": "model-00001-of-00007.safetensors"
+  }
+}