Upload folder using huggingface_hub

config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "merged/ENDURANCE-DELLA",
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 12288,
+  "initializer_range": 0.02,
+  "intermediate_size": 28672,
+  "max_position_embeddings": 131072,
+  "model_type": "mistral",
+  "num_attention_heads": 96,
+  "num_hidden_layers": 72,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.47.0",
+  "use_cache": true,
+  "vocab_size": 32768
+}

model-00001-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c26afec5f8a613a8ae834a227b9763f958bd934f77e1236a6f294d203f821070
+size 4378928504

model-00002-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9d2f8916fedb8ab6a631db888d8a5a30202e2f12e3035610702b58af32c358a
+size 4907411088

model-00003-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92bee0c66fb80fe030cb2ffe8629f3694b0e5ee00888547a9e5b12b6c6516d73
+size 4806747904

model-00004-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cca1dcea77d15a99b61c83fdcddcfa26477b9e95710a9a369577bf657fb7222
+size 4831938544

model-00005-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:404ff2be54312c82a51197fe41dff3c1a6ef35013ba9a6d6ad2c391e533fc1ee
+size 4831938552

model-00006-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4702950309b0d0928a2f38dfa1c524e90391576ef78c019b5214aa1f315a5fad
+size 4907411096

model-00007-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2789b65ea69bf48411bdd361490b4be1bddbfbea74d66a2d694249d818eb1571
+size 4806747904

model-00008-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf8c4d5dd4804870339e534b94c06d06def96e598f8471a88b38e3bbfc3665bb
+size 4831938536

model-00009-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:600421ca5331c7a1af45797e7efa14983bf1321fdad26981246d825fc378f1da
+size 4831938552

model-00010-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92cfe7c4e5b4852af1e71e69a919aec931dc56065a4842d7fe38c5b238e3c3cf
+size 4907411096

model-00011-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7787d6b6e6f545e4f70d636c553879b8dcd565090b4e2115ce89f93b8a3b6ad3
+size 4806747904

model-00012-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df0548c45e1da018441cd71ee36106fd28e030fea638598a216a7eff97fa02cb
+size 4831938544

model-00013-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38fec4815674aaddeff968acde65435d08ce69c7d72ec3448bef1e8f88a987ee
+size 4831938552

model-00014-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c48df2053422df98a44c9b50630c735386fca57552942e37f3dcf29ae7c2020f
+size 4907411088

model-00015-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccee02b0e931e63021f537c7cafd9eaa1253f1e8a3b5f469fd2072a339094337
+size 4806747904

model-00016-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bc4e88736072427e68bd797693ff1ee6addc467f99e2c2e71d359802778d72e
+size 4831938544

model-00017-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abcafa630d3a8cd5edd0c6273157b8b21e2ae36e6c1bc8527e80e2fc93bfffb0
+size 4831938552

model-00018-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a15f8d8c5435eaf1707db86207667c2ba56d7a8126ae4d3f260d06a58525df3
+size 4907411096

model-00019-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d04c2e6381a54c9ee66a25c11b6314b78150ce518302ef20a7e282d56c7b0db9
+size 4806747904

model-00020-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a114cf24d96842509918b98e15d7acdbb6db5ad7721c7a112d164213921913ef
+size 4831938544

model-00021-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4f2b82d9229f6d883ce49e879c1ad731ad42f0ec3e8683436b1f35e8b9239b1
+size 4831938544

model-00022-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd7e131d6b4a24035c9668d43d54c89b03816af61db17f8d713952860c1d979e
+size 4907411096

model-00023-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6a40ed9bb900643d3c1ef9e7dfc24a353465d3b3fea44b9389a90a2bdd3649d
+size 4806747904

model-00024-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc788c348c85213d164f73cd8d911d0a7f8c5b809ab3d3389b36406df9e7e1bb
+size 4831938544

model-00025-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba4342205e9d766cd98ce4e649cd47c2680c69b16f17c8b5992e1e31a84a042d
+size 4831938552

model-00026-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c44a0e8e078d5d5ce680e0b6e23a0ada7411967f3ec7c72c9423041ca443aa5
+size 4907411096

model-00027-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1782295acbcc4b8687dfe537e3be9ca0fe823ef09e9f590199982c2a2f0c67b2
+size 4806747896

model-00028-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff9b8a0b7d628fee96fb2cebf0002a7cf01b12947000f2a7d48b891eda33a103
+size 4831938544

model-00029-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64a2356801ecd9dfd8e9db2071c23e15b48537e8ef622e3ecced5a0ad0caf1f7
+size 4831938552

model-00030-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54f02332e5af39e0a04bdde85cafc4ef5d4b9f30275b2c92426c4f5c98a38c85
+size 4907411096

model-00031-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a32d122013996428381e531a1b4c8b0e3735d43a226743136ec7c90b9527db3e
+size 4806747904

model-00032-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96e2c9812e22c9aed24cd3f9374fb93922cf5bec40122176d8c4289bb486ea9f
+size 4831938544

model-00033-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b16c6af207dc1ee696aa92dd3b1a4f8ac3f6b7c15fc69c11c305d8572c8beae
+size 4831938544

model-00034-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3825a75577d7cfaffbb1410a01ffc3cd5dea684803fba4a8ad72d66c82946c8a
+size 4907411096

model-00035-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b7c1f2a9c2f1888cafa23853eecd1288dbfcc0a80a28b2be32d8e3e9d48e8f1
+size 4806747904

model-00036-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:caa819146a620f7617a57cc8385ba7a97aacb1e327defcaa1d130572384f363b
+size 4831938544

model-00037-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:071713cd0f749ea02c2857fabd02125dabe2564be1af4b0ec16c58f2a0d00b73
+size 4831938552

model-00038-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:986f10f5648825cc28dc16a8f92ee22c4368f1e107164fae5c58e4e12bf114a0
+size 4907411096

model-00039-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:878490dfe01256095db5d588562eed7d00e528782dcc76523cb2ec48d6b22691
+size 4806747904

model-00040-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0c5c1d1bb6fa21a2b9b7afc9a235b331400da4227001c6151f63594ef6ded8d
+size 4831938544

model-00041-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1de06de40e72055666c3abb800af9b69a4240e141cd5158421810fd3cbc88179
+size 4831938544

model-00042-of-00042.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4a451b3e93990e31bc922a810c21bff93bfaa3fd471c4f58be9f38e196f5fff
+size 2768290856

model.safetensors.index.json

@@ -0,0 +1 @@
+{"metadata": {"mergekit_version": "0.0.5.2", "total_size": 200927502336}, "weight_map": {"lm_head.weight": "model-00001-of-00042.safetensors", "model.embed_tokens.weight": "model-00001-of-00042.safetensors", "model.layers.0.input_layernorm.weight": "model-00001-of-00042.safetensors", "model.layers.0.mlp.down_proj.weight": "model-00001-of-00042.safetensors", "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00042.safetensors", "model.layers.0.mlp.up_proj.weight": "model-00001-of-00042.safetensors", "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00042.safetensors", "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00042.safetensors", "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00042.safetensors", "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00042.safetensors", "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00042.safetensors", "model.layers.1.input_layernorm.weight": "model-00001-of-00042.safetensors", "model.layers.1.mlp.down_proj.weight": "model-00002-of-00042.safetensors", "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00042.safetensors", "model.layers.1.mlp.up_proj.weight": "model-00002-of-00042.safetensors", "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00042.safetensors", "model.layers.1.self_attn.k_proj.weight": "model-00002-of-00042.safetensors", "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00042.safetensors", "model.layers.1.self_attn.q_proj.weight": "model-00002-of-00042.safetensors", "model.layers.1.self_attn.v_proj.weight": "model-00002-of-00042.safetensors", "model.layers.10.input_layernorm.weight": "model-00002-of-00042.safetensors", "model.layers.10.mlp.down_proj.weight": "model-00002-of-00042.safetensors", "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00042.safetensors", "model.layers.10.mlp.up_proj.weight": "model-00002-of-00042.safetensors", "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00042.safetensors", "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00042.safetensors", "model.layers.10.self_attn.o_proj.weight": "model-00003-of-00042.safetensors", "model.layers.10.self_attn.q_proj.weight": "model-00003-of-00042.safetensors", "model.layers.10.self_attn.v_proj.weight": "model-00003-of-00042.safetensors", "model.layers.11.input_layernorm.weight": "model-00003-of-00042.safetensors", "model.layers.11.mlp.down_proj.weight": "model-00003-of-00042.safetensors", "model.layers.11.mlp.gate_proj.weight": "model-00003-of-00042.safetensors", "model.layers.11.mlp.up_proj.weight": "model-00003-of-00042.safetensors", "model.layers.11.post_attention_layernorm.weight": "model-00003-of-00042.safetensors", "model.layers.11.self_attn.k_proj.weight": "model-00003-of-00042.safetensors", "model.layers.11.self_attn.o_proj.weight": "model-00003-of-00042.safetensors", "model.layers.11.self_attn.q_proj.weight": "model-00003-of-00042.safetensors", "model.layers.11.self_attn.v_proj.weight": "model-00003-of-00042.safetensors", "model.layers.12.input_layernorm.weight": "model-00003-of-00042.safetensors", "model.layers.12.mlp.down_proj.weight": "model-00003-of-00042.safetensors", "model.layers.12.mlp.gate_proj.weight": "model-00003-of-00042.safetensors", "model.layers.12.mlp.up_proj.weight": "model-00004-of-00042.safetensors", "model.layers.12.post_attention_layernorm.weight": "model-00004-of-00042.safetensors", "model.layers.12.self_attn.k_proj.weight": "model-00004-of-00042.safetensors", "model.layers.12.self_attn.o_proj.weight": "model-00004-of-00042.safetensors", "model.layers.12.self_attn.q_proj.weight": "model-00004-of-00042.safetensors", "model.layers.12.self_attn.v_proj.weight": "model-00004-of-00042.safetensors", "model.layers.13.input_layernorm.weight": "model-00004-of-00042.safetensors", "model.layers.13.mlp.down_proj.weight": "model-00004-of-00042.safetensors", "model.layers.13.mlp.gate_proj.weight": "model-00004-of-00042.safetensors", "model.layers.13.mlp.up_proj.weight": "model-00004-of-00042.safetensors", "model.layers.13.post_attention_layernorm.weight": "model-00004-of-00042.safetensors", "model.layers.13.self_attn.k_proj.weight": "model-00004-of-00042.safetensors", "model.layers.13.self_attn.o_proj.weight": "model-00004-of-00042.safetensors", "model.layers.13.self_attn.q_proj.weight": "model-00004-of-00042.safetensors", "model.layers.13.self_attn.v_proj.weight": "model-00004-of-00042.safetensors", "model.layers.14.input_layernorm.weight": "model-00004-of-00042.safetensors", "model.layers.14.mlp.down_proj.weight": "model-00004-of-00042.safetensors", "model.layers.14.mlp.gate_proj.weight": "model-00005-of-00042.safetensors", "model.layers.14.mlp.up_proj.weight": "model-00005-of-00042.safetensors", "model.layers.14.post_attention_layernorm.weight": "model-00005-of-00042.safetensors", "model.layers.14.self_attn.k_proj.weight": "model-00005-of-00042.safetensors", "model.layers.14.self_attn.o_proj.weight": "model-00005-of-00042.safetensors", "model.layers.14.self_attn.q_proj.weight": "model-00005-of-00042.safetensors", "model.layers.14.self_attn.v_proj.weight": "model-00005-of-00042.safetensors", "model.layers.15.input_layernorm.weight": "model-00005-of-00042.safetensors", "model.layers.15.mlp.down_proj.weight": "model-00005-of-00042.safetensors", "model.layers.15.mlp.gate_proj.weight": "model-00005-of-00042.safetensors", "model.layers.15.mlp.up_proj.weight": "model-00005-of-00042.safetensors", "model.layers.15.post_attention_layernorm.weight": "model-00005-of-00042.safetensors", "model.layers.15.self_attn.k_proj.weight": "model-00005-of-00042.safetensors", "model.layers.15.self_attn.o_proj.weight": "model-00005-of-00042.safetensors", "model.layers.15.self_attn.q_proj.weight": "model-00005-of-00042.safetensors", "model.layers.15.self_attn.v_proj.weight": "model-00005-of-00042.safetensors", "model.layers.16.input_layernorm.weight": "model-00005-of-00042.safetensors", "model.layers.16.mlp.down_proj.weight": "model-00006-of-00042.safetensors", "model.layers.16.mlp.gate_proj.weight": "model-00006-of-00042.safetensors", "model.layers.16.mlp.up_proj.weight": "model-00006-of-00042.safetensors", "model.layers.16.post_attention_layernorm.weight": "model-00006-of-00042.safetensors", "model.layers.16.self_attn.k_proj.weight": "model-00006-of-00042.safetensors", "model.layers.16.self_attn.o_proj.weight": "model-00006-of-00042.safetensors", "model.layers.16.self_attn.q_proj.weight": "model-00006-of-00042.safetensors", "model.layers.16.self_attn.v_proj.weight": "model-00006-of-00042.safetensors", "model.layers.17.input_layernorm.weight": "model-00006-of-00042.safetensors", "model.layers.17.mlp.down_proj.weight": "model-00006-of-00042.safetensors", "model.layers.17.mlp.gate_proj.weight": "model-00006-of-00042.safetensors", "model.layers.17.mlp.up_proj.weight": "model-00006-of-00042.safetensors", "model.layers.17.post_attention_layernorm.weight": "model-00006-of-00042.safetensors", "model.layers.17.self_attn.k_proj.weight": "model-00006-of-00042.safetensors", "model.layers.17.self_attn.o_proj.weight": "model-00007-of-00042.safetensors", "model.layers.17.self_attn.q_proj.weight": "model-00007-of-00042.safetensors", "model.layers.17.self_attn.v_proj.weight": "model-00007-of-00042.safetensors", "model.layers.18.input_layernorm.weight": "model-00007-of-00042.safetensors", "model.layers.18.mlp.down_proj.weight": "model-00007-of-00042.safetensors", "model.layers.18.mlp.gate_proj.weight": "model-00007-of-00042.safetensors", "model.layers.18.mlp.up_proj.weight": "model-00007-of-00042.safetensors", "model.layers.18.post_attention_layernorm.weight": "model-00007-of-00042.safetensors", "model.layers.18.self_attn.k_proj.weight": "model-00007-of-00042.safetensors", "model.layers.18.self_attn.o_proj.weight": "model-00007-of-00042.safetensors", "model.layers.18.self_attn.q_proj.weight": "model-00007-of-00042.safetensors", "model.layers.18.self_attn.v_proj.weight": "model-00007-of-00042.safetensors", "model.layers.19.input_layernorm.weight": "model-00007-of-00042.safetensors", "model.layers.19.mlp.down_proj.weight": "model-00007-of-00042.safetensors", "model.layers.19.mlp.gate_proj.weight": "model-00007-of-00042.safetensors", "model.layers.19.mlp.up_proj.weight": "model-00008-of-00042.safetensors", "model.layers.19.post_attention_layernorm.weight": "model-00008-of-00042.safetensors", "model.layers.19.self_attn.k_proj.weight": "model-00008-of-00042.safetensors", "model.layers.19.self_attn.o_proj.weight": "model-00008-of-00042.safetensors", "model.layers.19.self_attn.q_proj.weight": "model-00008-of-00042.safetensors", "model.layers.19.self_attn.v_proj.weight": "model-00008-of-00042.safetensors", "model.layers.2.input_layernorm.weight": "model-00008-of-00042.safetensors", "model.layers.2.mlp.down_proj.weight": "model-00008-of-00042.safetensors", "model.layers.2.mlp.gate_proj.weight": "model-00008-of-00042.safetensors", "model.layers.2.mlp.up_proj.weight": "model-00008-of-00042.safetensors", "model.layers.2.post_attention_layernorm.weight": "model-00008-of-00042.safetensors", "model.layers.2.self_attn.k_proj.weight": "model-00008-of-00042.safetensors", "model.layers.2.self_attn.o_proj.weight": "model-00008-of-00042.safetensors", "model.layers.2.self_attn.q_proj.weight": "model-00008-of-00042.safetensors", "model.layers.2.self_attn.v_proj.weight": "model-00008-of-00042.safetensors", "model.layers.20.input_layernorm.weight": "model-00008-of-00042.safetensors", "model.layers.20.mlp.down_proj.weight": "model-00008-of-00042.safetensors", "model.layers.20.mlp.gate_proj.weight": "model-00009-of-00042.safetensors", "model.layers.20.mlp.up_proj.weight": "model-00009-of-00042.safetensors", "model.layers.20.post_attention_layernorm.weight": "model-00009-of-00042.safetensors", "model.layers.20.self_attn.k_proj.weight": "model-00009-of-00042.safetensors", "model.layers.20.self_attn.o_proj.weight": "model-00009-of-00042.safetensors", "model.layers.20.self_attn.q_proj.weight": "model-00009-of-00042.safetensors", "model.layers.20.self_attn.v_proj.weight": "model-00009-of-00042.safetensors", "model.layers.21.input_layernorm.weight": "model-00009-of-00042.safetensors", "model.layers.21.mlp.down_proj.weight": "model-00009-of-00042.safetensors", "model.layers.21.mlp.gate_proj.weight": "model-00009-of-00042.safetensors", "model.layers.21.mlp.up_proj.weight": "model-00009-of-00042.safetensors", "model.layers.21.post_attention_layernorm.weight": "model-00009-of-00042.safetensors", "model.layers.21.self_attn.k_proj.weight": "model-00009-of-00042.safetensors", "model.layers.21.self_attn.o_proj.weight": "model-00009-of-00042.safetensors", "model.layers.21.self_attn.q_proj.weight": "model-00009-of-00042.safetensors", "model.layers.21.self_attn.v_proj.weight": "model-00009-of-00042.safetensors", "model.layers.22.input_layernorm.weight": "model-00009-of-00042.safetensors", "model.layers.22.mlp.down_proj.weight": "model-00010-of-00042.safetensors", "model.layers.22.mlp.gate_proj.weight": "model-00010-of-00042.safetensors", "model.layers.22.mlp.up_proj.weight": "model-00010-of-00042.safetensors", "model.layers.22.post_attention_layernorm.weight": "model-00010-of-00042.safetensors", "model.layers.22.self_attn.k_proj.weight": "model-00010-of-00042.safetensors", "model.layers.22.self_attn.o_proj.weight": "model-00010-of-00042.safetensors", "model.layers.22.self_attn.q_proj.weight": "model-00010-of-00042.safetensors", "model.layers.22.self_attn.v_proj.weight": "model-00010-of-00042.safetensors", "model.layers.23.input_layernorm.weight": "model-00010-of-00042.safetensors", "model.layers.23.mlp.down_proj.weight": "model-00010-of-00042.safetensors", "model.layers.23.mlp.gate_proj.weight": "model-00010-of-00042.safetensors", "model.layers.23.mlp.up_proj.weight": "model-00010-of-00042.safetensors", "model.layers.23.post_attention_layernorm.weight": "model-00010-of-00042.safetensors", "model.layers.23.self_attn.k_proj.weight": "model-00010-of-00042.safetensors", "model.layers.23.self_attn.o_proj.weight": "model-00011-of-00042.safetensors", "model.layers.23.self_attn.q_proj.weight": "model-00011-of-00042.safetensors", "model.layers.23.self_attn.v_proj.weight": "model-00011-of-00042.safetensors", "model.layers.24.input_layernorm.weight": "model-00011-of-00042.safetensors", "model.layers.24.mlp.down_proj.weight": "model-00011-of-00042.safetensors", "model.layers.24.mlp.gate_proj.weight": "model-00011-of-00042.safetensors", "model.layers.24.mlp.up_proj.weight": "model-00011-of-00042.safetensors", "model.layers.24.post_attention_layernorm.weight": "model-00011-of-00042.safetensors", "model.layers.24.self_attn.k_proj.weight": "model-00011-of-00042.safetensors", "model.layers.24.self_attn.o_proj.weight": "model-00011-of-00042.safetensors", "model.layers.24.self_attn.q_proj.weight": "model-00011-of-00042.safetensors", "model.layers.24.self_attn.v_proj.weight": "model-00011-of-00042.safetensors", "model.layers.25.input_layernorm.weight": "model-00011-of-00042.safetensors", "model.layers.25.mlp.down_proj.weight": "model-00011-of-00042.safetensors", "model.layers.25.mlp.gate_proj.weight": "model-00011-of-00042.safetensors", "model.layers.25.mlp.up_proj.weight": "model-00012-of-00042.safetensors", "model.layers.25.post_attention_layernorm.weight": "model-00012-of-00042.safetensors", "model.layers.25.self_attn.k_proj.weight": "model-00012-of-00042.safetensors", "model.layers.25.self_attn.o_proj.weight": "model-00012-of-00042.safetensors", "model.layers.25.self_attn.q_proj.weight": "model-00012-of-00042.safetensors", "model.layers.25.self_attn.v_proj.weight": "model-00012-of-00042.safetensors", "model.layers.26.input_layernorm.weight": "model-00012-of-00042.safetensors", "model.layers.26.mlp.down_proj.weight": "model-00012-of-00042.safetensors", "model.layers.26.mlp.gate_proj.weight": "model-00012-of-00042.safetensors", "model.layers.26.mlp.up_proj.weight": "model-00012-of-00042.safetensors", "model.layers.26.post_attention_layernorm.weight": "model-00012-of-00042.safetensors", "model.layers.26.self_attn.k_proj.weight": "model-00012-of-00042.safetensors", "model.layers.26.self_attn.o_proj.weight": "model-00012-of-00042.safetensors", "model.layers.26.self_attn.q_proj.weight": "model-00012-of-00042.safetensors", "model.layers.26.self_attn.v_proj.weight": "model-00012-of-00042.safetensors", "model.layers.27.input_layernorm.weight": "model-00012-of-00042.safetensors", "model.layers.27.mlp.down_proj.weight": "model-00012-of-00042.safetensors", "model.layers.27.mlp.gate_proj.weight": "model-00013-of-00042.safetensors", "model.layers.27.mlp.up_proj.weight": "model-00013-of-00042.safetensors", "model.layers.27.post_attention_layernorm.weight": "model-00013-of-00042.safetensors", "model.layers.27.self_attn.k_proj.weight": "model-00013-of-00042.safetensors", "model.layers.27.self_attn.o_proj.weight": "model-00013-of-00042.safetensors", "model.layers.27.self_attn.q_proj.weight": "model-00013-of-00042.safetensors", "model.layers.27.self_attn.v_proj.weight": "model-00013-of-00042.safetensors", "model.layers.28.input_layernorm.weight": "model-00013-of-00042.safetensors", "model.layers.28.mlp.down_proj.weight": "model-00013-of-00042.safetensors", "model.layers.28.mlp.gate_proj.weight": "model-00013-of-00042.safetensors", "model.layers.28.mlp.up_proj.weight": "model-00013-of-00042.safetensors", "model.layers.28.post_attention_layernorm.weight": "model-00013-of-00042.safetensors", "model.layers.28.self_attn.k_proj.weight": "model-00013-of-00042.safetensors", "model.layers.28.self_attn.o_proj.weight": "model-00013-of-00042.safetensors", "model.layers.28.self_attn.q_proj.weight": "model-00013-of-00042.safetensors", "model.layers.28.self_attn.v_proj.weight": "model-00013-of-00042.safetensors", "model.layers.29.input_layernorm.weight": "model-00013-of-00042.safetensors", "model.layers.29.mlp.down_proj.weight": "model-00014-of-00042.safetensors", "model.layers.29.mlp.gate_proj.weight": "model-00014-of-00042.safetensors", "model.layers.29.mlp.up_proj.weight": "model-00014-of-00042.safetensors", "model.layers.29.post_attention_layernorm.weight": "model-00014-of-00042.safetensors", "model.layers.29.self_attn.k_proj.weight": "model-00014-of-00042.safetensors", "model.layers.29.self_attn.o_proj.weight": "model-00014-of-00042.safetensors", "model.layers.29.self_attn.q_proj.weight": "model-00014-of-00042.safetensors", "model.layers.29.self_attn.v_proj.weight": "model-00014-of-00042.safetensors", "model.layers.3.input_layernorm.weight": "model-00014-of-00042.safetensors", "model.layers.3.mlp.down_proj.weight": "model-00014-of-00042.safetensors", "model.layers.3.mlp.gate_proj.weight": "model-00014-of-00042.safetensors", "model.layers.3.mlp.up_proj.weight": "model-00014-of-00042.safetensors", "model.layers.3.post_attention_layernorm.weight": "model-00014-of-00042.safetensors", "model.layers.3.self_attn.k_proj.weight": "model-00014-of-00042.safetensors", "model.layers.3.self_attn.o_proj.weight": "model-00015-of-00042.safetensors", "model.layers.3.self_attn.q_proj.weight": "model-00015-of-00042.safetensors", "model.layers.3.self_attn.v_proj.weight": "model-00015-of-00042.safetensors", "model.layers.30.input_layernorm.weight": "model-00015-of-00042.safetensors", "model.layers.30.mlp.down_proj.weight": "model-00015-of-00042.safetensors", "model.layers.30.mlp.gate_proj.weight": "model-00015-of-00042.safetensors", "model.layers.30.mlp.up_proj.weight": "model-00015-of-00042.safetensors", "model.layers.30.post_attention_layernorm.weight": "model-00015-of-00042.safetensors", "model.layers.30.self_attn.k_proj.weight": "model-00015-of-00042.safetensors", "model.layers.30.self_attn.o_proj.weight": "model-00015-of-00042.safetensors", "model.layers.30.self_attn.q_proj.weight": "model-00015-of-00042.safetensors", "model.layers.30.self_attn.v_proj.weight": "model-00015-of-00042.safetensors", "model.layers.31.input_layernorm.weight": "model-00015-of-00042.safetensors", "model.layers.31.mlp.down_proj.weight": "model-00015-of-00042.safetensors", "model.layers.31.mlp.gate_proj.weight": "model-00015-of-00042.safetensors", "model.layers.31.mlp.up_proj.weight": "model-00016-of-00042.safetensors", "model.layers.31.post_attention_layernorm.weight": "model-00016-of-00042.safetensors", "model.layers.31.self_attn.k_proj.weight": "model-00016-of-00042.safetensors", "model.layers.31.self_attn.o_proj.weight": "model-00016-of-00042.safetensors", "model.layers.31.self_attn.q_proj.weight": "model-00016-of-00042.safetensors", "model.layers.31.self_attn.v_proj.weight": "model-00016-of-00042.safetensors", "model.layers.32.input_layernorm.weight": "model-00016-of-00042.safetensors", "model.layers.32.mlp.down_proj.weight": "model-00016-of-00042.safetensors", "model.layers.32.mlp.gate_proj.weight": "model-00016-of-00042.safetensors", "model.layers.32.mlp.up_proj.weight": "model-00016-of-00042.safetensors", "model.layers.32.post_attention_layernorm.weight": "model-00016-of-00042.safetensors", "model.layers.32.self_attn.k_proj.weight": "model-00016-of-00042.safetensors", "model.layers.32.self_attn.o_proj.weight": "model-00016-of-00042.safetensors", "model.layers.32.self_attn.q_proj.weight": "model-00016-of-00042.safetensors", "model.layers.32.self_attn.v_proj.weight": "model-00016-of-00042.safetensors", "model.layers.33.input_layernorm.weight": "model-00016-of-00042.safetensors", "model.layers.33.mlp.down_proj.weight": "model-00016-of-00042.safetensors", "model.layers.33.mlp.gate_proj.weight": "model-00017-of-00042.safetensors", "model.layers.33.mlp.up_proj.weight": "model-00017-of-00042.safetensors", "model.layers.33.post_attention_layernorm.weight": "model-00017-of-00042.safetensors", "model.layers.33.self_attn.k_proj.weight": "model-00017-of-00042.safetensors", "model.layers.33.self_attn.o_proj.weight": "model-00017-of-00042.safetensors", "model.layers.33.self_attn.q_proj.weight": "model-00017-of-00042.safetensors", "model.layers.33.self_attn.v_proj.weight": "model-00017-of-00042.safetensors", "model.layers.34.input_layernorm.weight": "model-00017-of-00042.safetensors", "model.layers.34.mlp.down_proj.weight": "model-00017-of-00042.safetensors", "model.layers.34.mlp.gate_proj.weight": "model-00017-of-00042.safetensors", "model.layers.34.mlp.up_proj.weight": "model-00017-of-00042.safetensors", "model.layers.34.post_attention_layernorm.weight": "model-00017-of-00042.safetensors", "model.layers.34.self_attn.k_proj.weight": "model-00017-of-00042.safetensors", "model.layers.34.self_attn.o_proj.weight": "model-00017-of-00042.safetensors", "model.layers.34.self_attn.q_proj.weight": "model-00017-of-00042.safetensors", "model.layers.34.self_attn.v_proj.weight": "model-00017-of-00042.safetensors", "model.layers.35.input_layernorm.weight": "model-00017-of-00042.safetensors", "model.layers.35.mlp.down_proj.weight": "model-00018-of-00042.safetensors", "model.layers.35.mlp.gate_proj.weight": "model-00018-of-00042.safetensors", "model.layers.35.mlp.up_proj.weight": "model-00018-of-00042.safetensors", "model.layers.35.post_attention_layernorm.weight": "model-00018-of-00042.safetensors", "model.layers.35.self_attn.k_proj.weight": "model-00018-of-00042.safetensors", "model.layers.35.self_attn.o_proj.weight": "model-00018-of-00042.safetensors", "model.layers.35.self_attn.q_proj.weight": "model-00018-of-00042.safetensors", "model.layers.35.self_attn.v_proj.weight": "model-00018-of-00042.safetensors", "model.layers.36.input_layernorm.weight": "model-00018-of-00042.safetensors", "model.layers.36.mlp.down_proj.weight": "model-00018-of-00042.safetensors", "model.layers.36.mlp.gate_proj.weight": "model-00018-of-00042.safetensors", "model.layers.36.mlp.up_proj.weight": "model-00018-of-00042.safetensors", "model.layers.36.post_attention_layernorm.weight": "model-00018-of-00042.safetensors", "model.layers.36.self_attn.k_proj.weight": "model-00018-of-00042.safetensors", "model.layers.36.self_attn.o_proj.weight": "model-00019-of-00042.safetensors", "model.layers.36.self_attn.q_proj.weight": "model-00019-of-00042.safetensors", "model.layers.36.self_attn.v_proj.weight": "model-00019-of-00042.safetensors", "model.layers.37.input_layernorm.weight": "model-00019-of-00042.safetensors", "model.layers.37.mlp.down_proj.weight": "model-00019-of-00042.safetensors", "model.layers.37.mlp.gate_proj.weight": "model-00019-of-00042.safetensors", "model.layers.37.mlp.up_proj.weight": "model-00019-of-00042.safetensors", "model.layers.37.post_attention_layernorm.weight": "model-00019-of-00042.safetensors", "model.layers.37.self_attn.k_proj.weight": "model-00019-of-00042.safetensors", "model.layers.37.self_attn.o_proj.weight": "model-00019-of-00042.safetensors", "model.layers.37.self_attn.q_proj.weight": "model-00019-of-00042.safetensors", "model.layers.37.self_attn.v_proj.weight": "model-00019-of-00042.safetensors", "model.layers.38.input_layernorm.weight": "model-00019-of-00042.safetensors", "model.layers.38.mlp.down_proj.weight": "model-00019-of-00042.safetensors", "model.layers.38.mlp.gate_proj.weight": "model-00019-of-00042.safetensors", "model.layers.38.mlp.up_proj.weight": "model-00020-of-00042.safetensors", "model.layers.38.post_attention_layernorm.weight": "model-00020-of-00042.safetensors", "model.layers.38.self_attn.k_proj.weight": "model-00020-of-00042.safetensors", "model.layers.38.self_attn.o_proj.weight": "model-00020-of-00042.safetensors", "model.layers.38.self_attn.q_proj.weight": "model-00020-of-00042.safetensors", "model.layers.38.self_attn.v_proj.weight": "model-00020-of-00042.safetensors", "model.layers.39.input_layernorm.weight": "model-00020-of-00042.safetensors", "model.layers.39.mlp.down_proj.weight": "model-00020-of-00042.safetensors", "model.layers.39.mlp.gate_proj.weight": "model-00020-of-00042.safetensors", "model.layers.39.mlp.up_proj.weight": "model-00020-of-00042.safetensors", "model.layers.39.post_attention_layernorm.weight": "model-00020-of-00042.safetensors", "model.layers.39.self_attn.k_proj.weight": "model-00020-of-00042.safetensors", "model.layers.39.self_attn.o_proj.weight": "model-00020-of-00042.safetensors", "model.layers.39.self_attn.q_proj.weight": "model-00020-of-00042.safetensors", "model.layers.39.self_attn.v_proj.weight": "model-00020-of-00042.safetensors", "model.layers.4.input_layernorm.weight": "model-00020-of-00042.safetensors", "model.layers.4.mlp.down_proj.weight": "model-00020-of-00042.safetensors", "model.layers.4.mlp.gate_proj.weight": "model-00021-of-00042.safetensors", "model.layers.4.mlp.up_proj.weight": "model-00021-of-00042.safetensors", "model.layers.4.post_attention_layernorm.weight": "model-00021-of-00042.safetensors", "model.layers.4.self_attn.k_proj.weight": "model-00021-of-00042.safetensors", "model.layers.4.self_attn.o_proj.weight": "model-00021-of-00042.safetensors", "model.layers.4.self_attn.q_proj.weight": "model-00021-of-00042.safetensors", "model.layers.4.self_attn.v_proj.weight": "model-00021-of-00042.safetensors", "model.layers.40.input_layernorm.weight": "model-00021-of-00042.safetensors", "model.layers.40.mlp.down_proj.weight": "model-00021-of-00042.safetensors", "model.layers.40.mlp.gate_proj.weight": "model-00021-of-00042.safetensors", "model.layers.40.mlp.up_proj.weight": "model-00021-of-00042.safetensors", "model.layers.40.post_attention_layernorm.weight": "model-00021-of-00042.safetensors", "model.layers.40.self_attn.k_proj.weight": "model-00021-of-00042.safetensors", "model.layers.40.self_attn.o_proj.weight": "model-00021-of-00042.safetensors", "model.layers.40.self_attn.q_proj.weight": "model-00021-of-00042.safetensors", "model.layers.40.self_attn.v_proj.weight": "model-00021-of-00042.safetensors", "model.layers.41.input_layernorm.weight": "model-00021-of-00042.safetensors", "model.layers.41.mlp.down_proj.weight": "model-00022-of-00042.safetensors", "model.layers.41.mlp.gate_proj.weight": "model-00022-of-00042.safetensors", "model.layers.41.mlp.up_proj.weight": "model-00022-of-00042.safetensors", "model.layers.41.post_attention_layernorm.weight": "model-00022-of-00042.safetensors", "model.layers.41.self_attn.k_proj.weight": "model-00022-of-00042.safetensors", "model.layers.41.self_attn.o_proj.weight": "model-00022-of-00042.safetensors", "model.layers.41.self_attn.q_proj.weight": "model-00022-of-00042.safetensors", "model.layers.41.self_attn.v_proj.weight": "model-00022-of-00042.safetensors", "model.layers.42.input_layernorm.weight": "model-00022-of-00042.safetensors", "model.layers.42.mlp.down_proj.weight": "model-00022-of-00042.safetensors", "model.layers.42.mlp.gate_proj.weight": "model-00022-of-00042.safetensors", "model.layers.42.mlp.up_proj.weight": "model-00022-of-00042.safetensors", "model.layers.42.post_attention_layernorm.weight": "model-00022-of-00042.safetensors", "model.layers.42.self_attn.k_proj.weight": "model-00022-of-00042.safetensors", "model.layers.42.self_attn.o_proj.weight": "model-00023-of-00042.safetensors", "model.layers.42.self_attn.q_proj.weight": "model-00023-of-00042.safetensors", "model.layers.42.self_attn.v_proj.weight": "model-00023-of-00042.safetensors", "model.layers.43.input_layernorm.weight": "model-00023-of-00042.safetensors", "model.layers.43.mlp.down_proj.weight": "model-00023-of-00042.safetensors", "model.layers.43.mlp.gate_proj.weight": "model-00023-of-00042.safetensors", "model.layers.43.mlp.up_proj.weight": "model-00023-of-00042.safetensors", "model.layers.43.post_attention_layernorm.weight": "model-00023-of-00042.safetensors", "model.layers.43.self_attn.k_proj.weight": "model-00023-of-00042.safetensors", "model.layers.43.self_attn.o_proj.weight": "model-00023-of-00042.safetensors", "model.layers.43.self_attn.q_proj.weight": "model-00023-of-00042.safetensors", "model.layers.43.self_attn.v_proj.weight": "model-00023-of-00042.safetensors", "model.layers.44.input_layernorm.weight": "model-00023-of-00042.safetensors", "model.layers.44.mlp.down_proj.weight": "model-00023-of-00042.safetensors", "model.layers.44.mlp.gate_proj.weight": "model-00023-of-00042.safetensors", "model.layers.44.mlp.up_proj.weight": "model-00024-of-00042.safetensors", "model.layers.44.post_attention_layernorm.weight": "model-00024-of-00042.safetensors", "model.layers.44.self_attn.k_proj.weight": "model-00024-of-00042.safetensors", "model.layers.44.self_attn.o_proj.weight": "model-00024-of-00042.safetensors", "model.layers.44.self_attn.q_proj.weight": "model-00024-of-00042.safetensors", "model.layers.44.self_attn.v_proj.weight": "model-00024-of-00042.safetensors", "model.layers.45.input_layernorm.weight": "model-00024-of-00042.safetensors", "model.layers.45.mlp.down_proj.weight": "model-00024-of-00042.safetensors", "model.layers.45.mlp.gate_proj.weight": "model-00024-of-00042.safetensors", "model.layers.45.mlp.up_proj.weight": "model-00024-of-00042.safetensors", "model.layers.45.post_attention_layernorm.weight": "model-00024-of-00042.safetensors", "model.layers.45.self_attn.k_proj.weight": "model-00024-of-00042.safetensors", "model.layers.45.self_attn.o_proj.weight": "model-00024-of-00042.safetensors", "model.layers.45.self_attn.q_proj.weight": "model-00024-of-00042.safetensors", "model.layers.45.self_attn.v_proj.weight": "model-00024-of-00042.safetensors", "model.layers.46.input_layernorm.weight": "model-00024-of-00042.safetensors", "model.layers.46.mlp.down_proj.weight": "model-00024-of-00042.safetensors", "model.layers.46.mlp.gate_proj.weight": "model-00025-of-00042.safetensors", "model.layers.46.mlp.up_proj.weight": "model-00025-of-00042.safetensors", "model.layers.46.post_attention_layernorm.weight": "model-00025-of-00042.safetensors", "model.layers.46.self_attn.k_proj.weight": "model-00025-of-00042.safetensors", "model.layers.46.self_attn.o_proj.weight": "model-00025-of-00042.safetensors", "model.layers.46.self_attn.q_proj.weight": "model-00025-of-00042.safetensors", "model.layers.46.self_attn.v_proj.weight": "model-00025-of-00042.safetensors", "model.layers.47.input_layernorm.weight": "model-00025-of-00042.safetensors", "model.layers.47.mlp.down_proj.weight": "model-00025-of-00042.safetensors", "model.layers.47.mlp.gate_proj.weight": "model-00025-of-00042.safetensors", "model.layers.47.mlp.up_proj.weight": "model-00025-of-00042.safetensors", "model.layers.47.post_attention_layernorm.weight": "model-00025-of-00042.safetensors", "model.layers.47.self_attn.k_proj.weight": "model-00025-of-00042.safetensors", "model.layers.47.self_attn.o_proj.weight": "model-00025-of-00042.safetensors", "model.layers.47.self_attn.q_proj.weight": "model-00025-of-00042.safetensors", "model.layers.47.self_attn.v_proj.weight": "model-00025-of-00042.safetensors", "model.layers.48.input_layernorm.weight": "model-00025-of-00042.safetensors", "model.layers.48.mlp.down_proj.weight": "model-00026-of-00042.safetensors", "model.layers.48.mlp.gate_proj.weight": "model-00026-of-00042.safetensors", "model.layers.48.mlp.up_proj.weight": "model-00026-of-00042.safetensors", "model.layers.48.post_attention_layernorm.weight": "model-00026-of-00042.safetensors", "model.layers.48.self_attn.k_proj.weight": "model-00026-of-00042.safetensors", "model.layers.48.self_attn.o_proj.weight": "model-00026-of-00042.safetensors", "model.layers.48.self_attn.q_proj.weight": "model-00026-of-00042.safetensors", "model.layers.48.self_attn.v_proj.weight": "model-00026-of-00042.safetensors", "model.layers.49.input_layernorm.weight": "model-00026-of-00042.safetensors", "model.layers.49.mlp.down_proj.weight": "model-00026-of-00042.safetensors", "model.layers.49.mlp.gate_proj.weight": "model-00026-of-00042.safetensors", "model.layers.49.mlp.up_proj.weight": "model-00026-of-00042.safetensors", "model.layers.49.post_attention_layernorm.weight": "model-00026-of-00042.safetensors", "model.layers.49.self_attn.k_proj.weight": "model-00026-of-00042.safetensors", "model.layers.49.self_attn.o_proj.weight": "model-00027-of-00042.safetensors", "model.layers.49.self_attn.q_proj.weight": "model-00027-of-00042.safetensors", "model.layers.49.self_attn.v_proj.weight": "model-00027-of-00042.safetensors", "model.layers.5.input_layernorm.weight": "model-00027-of-00042.safetensors", "model.layers.5.mlp.down_proj.weight": "model-00027-of-00042.safetensors", "model.layers.5.mlp.gate_proj.weight": "model-00027-of-00042.safetensors", "model.layers.5.mlp.up_proj.weight": "model-00027-of-00042.safetensors", "model.layers.5.post_attention_layernorm.weight": "model-00027-of-00042.safetensors", "model.layers.5.self_attn.k_proj.weight": "model-00027-of-00042.safetensors", "model.layers.5.self_attn.o_proj.weight": "model-00027-of-00042.safetensors", "model.layers.5.self_attn.q_proj.weight": "model-00027-of-00042.safetensors", "model.layers.5.self_attn.v_proj.weight": "model-00027-of-00042.safetensors", "model.layers.50.input_layernorm.weight": "model-00027-of-00042.safetensors", "model.layers.50.mlp.down_proj.weight": "model-00027-of-00042.safetensors", "model.layers.50.mlp.gate_proj.weight": "model-00027-of-00042.safetensors", "model.layers.50.mlp.up_proj.weight": "model-00028-of-00042.safetensors", "model.layers.50.post_attention_layernorm.weight": "model-00028-of-00042.safetensors", "model.layers.50.self_attn.k_proj.weight": "model-00028-of-00042.safetensors", "model.layers.50.self_attn.o_proj.weight": "model-00028-of-00042.safetensors", "model.layers.50.self_attn.q_proj.weight": "model-00028-of-00042.safetensors", "model.layers.50.self_attn.v_proj.weight": "model-00028-of-00042.safetensors", "model.layers.51.input_layernorm.weight": "model-00028-of-00042.safetensors", "model.layers.51.mlp.down_proj.weight": "model-00028-of-00042.safetensors", "model.layers.51.mlp.gate_proj.weight": "model-00028-of-00042.safetensors", "model.layers.51.mlp.up_proj.weight": "model-00028-of-00042.safetensors", "model.layers.51.post_attention_layernorm.weight": "model-00028-of-00042.safetensors", "model.layers.51.self_attn.k_proj.weight": "model-00028-of-00042.safetensors", "model.layers.51.self_attn.o_proj.weight": "model-00028-of-00042.safetensors", "model.layers.51.self_attn.q_proj.weight": "model-00028-of-00042.safetensors", "model.layers.51.self_attn.v_proj.weight": "model-00028-of-00042.safetensors", "model.layers.52.input_layernorm.weight": "model-00028-of-00042.safetensors", "model.layers.52.mlp.down_proj.weight": "model-00028-of-00042.safetensors", "model.layers.52.mlp.gate_proj.weight": "model-00029-of-00042.safetensors", "model.layers.52.mlp.up_proj.weight": "model-00029-of-00042.safetensors", "model.layers.52.post_attention_layernorm.weight": "model-00029-of-00042.safetensors", "model.layers.52.self_attn.k_proj.weight": "model-00029-of-00042.safetensors", "model.layers.52.self_attn.o_proj.weight": "model-00029-of-00042.safetensors", "model.layers.52.self_attn.q_proj.weight": "model-00029-of-00042.safetensors", "model.layers.52.self_attn.v_proj.weight": "model-00029-of-00042.safetensors", "model.layers.53.input_layernorm.weight": "model-00029-of-00042.safetensors", "model.layers.53.mlp.down_proj.weight": "model-00029-of-00042.safetensors", "model.layers.53.mlp.gate_proj.weight": "model-00029-of-00042.safetensors", "model.layers.53.mlp.up_proj.weight": "model-00029-of-00042.safetensors", "model.layers.53.post_attention_layernorm.weight": "model-00029-of-00042.safetensors", "model.layers.53.self_attn.k_proj.weight": "model-00029-of-00042.safetensors", "model.layers.53.self_attn.o_proj.weight": "model-00029-of-00042.safetensors", "model.layers.53.self_attn.q_proj.weight": "model-00029-of-00042.safetensors", "model.layers.53.self_attn.v_proj.weight": "model-00029-of-00042.safetensors", "model.layers.54.input_layernorm.weight": "model-00029-of-00042.safetensors", "model.layers.54.mlp.down_proj.weight": "model-00030-of-00042.safetensors", "model.layers.54.mlp.gate_proj.weight": "model-00030-of-00042.safetensors", "model.layers.54.mlp.up_proj.weight": "model-00030-of-00042.safetensors", "model.layers.54.post_attention_layernorm.weight": "model-00030-of-00042.safetensors", "model.layers.54.self_attn.k_proj.weight": "model-00030-of-00042.safetensors", "model.layers.54.self_attn.o_proj.weight": "model-00030-of-00042.safetensors", "model.layers.54.self_attn.q_proj.weight": "model-00030-of-00042.safetensors", "model.layers.54.self_attn.v_proj.weight": "model-00030-of-00042.safetensors", "model.layers.55.input_layernorm.weight": "model-00030-of-00042.safetensors", "model.layers.55.mlp.down_proj.weight": "model-00030-of-00042.safetensors", "model.layers.55.mlp.gate_proj.weight": "model-00030-of-00042.safetensors", "model.layers.55.mlp.up_proj.weight": "model-00030-of-00042.safetensors", "model.layers.55.post_attention_layernorm.weight": "model-00030-of-00042.safetensors", "model.layers.55.self_attn.k_proj.weight": "model-00030-of-00042.safetensors", "model.layers.55.self_attn.o_proj.weight": "model-00031-of-00042.safetensors", "model.layers.55.self_attn.q_proj.weight": "model-00031-of-00042.safetensors", "model.layers.55.self_attn.v_proj.weight": "model-00031-of-00042.safetensors", "model.layers.56.input_layernorm.weight": "model-00031-of-00042.safetensors", "model.layers.56.mlp.down_proj.weight": "model-00031-of-00042.safetensors", "model.layers.56.mlp.gate_proj.weight": "model-00031-of-00042.safetensors", "model.layers.56.mlp.up_proj.weight": "model-00031-of-00042.safetensors", "model.layers.56.post_attention_layernorm.weight": "model-00031-of-00042.safetensors", "model.layers.56.self_attn.k_proj.weight": "model-00031-of-00042.safetensors", "model.layers.56.self_attn.o_proj.weight": "model-00031-of-00042.safetensors", "model.layers.56.self_attn.q_proj.weight": "model-00031-of-00042.safetensors", "model.layers.56.self_attn.v_proj.weight": "model-00031-of-00042.safetensors", "model.layers.57.input_layernorm.weight": "model-00031-of-00042.safetensors", "model.layers.57.mlp.down_proj.weight": "model-00031-of-00042.safetensors", "model.layers.57.mlp.gate_proj.weight": "model-00031-of-00042.safetensors", "model.layers.57.mlp.up_proj.weight": "model-00032-of-00042.safetensors", "model.layers.57.post_attention_layernorm.weight": "model-00032-of-00042.safetensors", "model.layers.57.self_attn.k_proj.weight": "model-00032-of-00042.safetensors", "model.layers.57.self_attn.o_proj.weight": "model-00032-of-00042.safetensors", "model.layers.57.self_attn.q_proj.weight": "model-00032-of-00042.safetensors", "model.layers.57.self_attn.v_proj.weight": "model-00032-of-00042.safetensors", "model.layers.58.input_layernorm.weight": "model-00032-of-00042.safetensors", "model.layers.58.mlp.down_proj.weight": "model-00032-of-00042.safetensors", "model.layers.58.mlp.gate_proj.weight": "model-00032-of-00042.safetensors", "model.layers.58.mlp.up_proj.weight": "model-00032-of-00042.safetensors", "model.layers.58.post_attention_layernorm.weight": "model-00032-of-00042.safetensors", "model.layers.58.self_attn.k_proj.weight": "model-00032-of-00042.safetensors", "model.layers.58.self_attn.o_proj.weight": "model-00032-of-00042.safetensors", "model.layers.58.self_attn.q_proj.weight": "model-00032-of-00042.safetensors", "model.layers.58.self_attn.v_proj.weight": "model-00032-of-00042.safetensors", "model.layers.59.input_layernorm.weight": "model-00032-of-00042.safetensors", "model.layers.59.mlp.down_proj.weight": "model-00032-of-00042.safetensors", "model.layers.59.mlp.gate_proj.weight": "model-00033-of-00042.safetensors", "model.layers.59.mlp.up_proj.weight": "model-00033-of-00042.safetensors", "model.layers.59.post_attention_layernorm.weight": "model-00033-of-00042.safetensors", "model.layers.59.self_attn.k_proj.weight": "model-00033-of-00042.safetensors", "model.layers.59.self_attn.o_proj.weight": "model-00033-of-00042.safetensors", "model.layers.59.self_attn.q_proj.weight": "model-00033-of-00042.safetensors", "model.layers.59.self_attn.v_proj.weight": "model-00033-of-00042.safetensors", "model.layers.6.input_layernorm.weight": "model-00033-of-00042.safetensors", "model.layers.6.mlp.down_proj.weight": "model-00033-of-00042.safetensors", "model.layers.6.mlp.gate_proj.weight": "model-00033-of-00042.safetensors", "model.layers.6.mlp.up_proj.weight": "model-00033-of-00042.safetensors", "model.layers.6.post_attention_layernorm.weight": "model-00033-of-00042.safetensors", "model.layers.6.self_attn.k_proj.weight": "model-00033-of-00042.safetensors", "model.layers.6.self_attn.o_proj.weight": "model-00033-of-00042.safetensors", "model.layers.6.self_attn.q_proj.weight": "model-00033-of-00042.safetensors", "model.layers.6.self_attn.v_proj.weight": "model-00033-of-00042.safetensors", "model.layers.60.input_layernorm.weight": "model-00033-of-00042.safetensors", "model.layers.60.mlp.down_proj.weight": "model-00034-of-00042.safetensors", "model.layers.60.mlp.gate_proj.weight": "model-00034-of-00042.safetensors", "model.layers.60.mlp.up_proj.weight": "model-00034-of-00042.safetensors", "model.layers.60.post_attention_layernorm.weight": "model-00034-of-00042.safetensors", "model.layers.60.self_attn.k_proj.weight": "model-00034-of-00042.safetensors", "model.layers.60.self_attn.o_proj.weight": "model-00034-of-00042.safetensors", "model.layers.60.self_attn.q_proj.weight": "model-00034-of-00042.safetensors", "model.layers.60.self_attn.v_proj.weight": "model-00034-of-00042.safetensors", "model.layers.61.input_layernorm.weight": "model-00034-of-00042.safetensors", "model.layers.61.mlp.down_proj.weight": "model-00034-of-00042.safetensors", "model.layers.61.mlp.gate_proj.weight": "model-00034-of-00042.safetensors", "model.layers.61.mlp.up_proj.weight": "model-00034-of-00042.safetensors", "model.layers.61.post_attention_layernorm.weight": "model-00034-of-00042.safetensors", "model.layers.61.self_attn.k_proj.weight": "model-00034-of-00042.safetensors", "model.layers.61.self_attn.o_proj.weight": "model-00035-of-00042.safetensors", "model.layers.61.self_attn.q_proj.weight": "model-00035-of-00042.safetensors", "model.layers.61.self_attn.v_proj.weight": "model-00035-of-00042.safetensors", "model.layers.62.input_layernorm.weight": "model-00035-of-00042.safetensors", "model.layers.62.mlp.down_proj.weight": "model-00035-of-00042.safetensors", "model.layers.62.mlp.gate_proj.weight": "model-00035-of-00042.safetensors", "model.layers.62.mlp.up_proj.weight": "model-00035-of-00042.safetensors", "model.layers.62.post_attention_layernorm.weight": "model-00035-of-00042.safetensors", "model.layers.62.self_attn.k_proj.weight": "model-00035-of-00042.safetensors", "model.layers.62.self_attn.o_proj.weight": "model-00035-of-00042.safetensors", "model.layers.62.self_attn.q_proj.weight": "model-00035-of-00042.safetensors", "model.layers.62.self_attn.v_proj.weight": "model-00035-of-00042.safetensors", "model.layers.63.input_layernorm.weight": "model-00035-of-00042.safetensors", "model.layers.63.mlp.down_proj.weight": "model-00035-of-00042.safetensors", "model.layers.63.mlp.gate_proj.weight": "model-00035-of-00042.safetensors", "model.layers.63.mlp.up_proj.weight": "model-00036-of-00042.safetensors", "model.layers.63.post_attention_layernorm.weight": "model-00036-of-00042.safetensors", "model.layers.63.self_attn.k_proj.weight": "model-00036-of-00042.safetensors", "model.layers.63.self_attn.o_proj.weight": "model-00036-of-00042.safetensors", "model.layers.63.self_attn.q_proj.weight": "model-00036-of-00042.safetensors", "model.layers.63.self_attn.v_proj.weight": "model-00036-of-00042.safetensors", "model.layers.64.input_layernorm.weight": "model-00036-of-00042.safetensors", "model.layers.64.mlp.down_proj.weight": "model-00036-of-00042.safetensors", "model.layers.64.mlp.gate_proj.weight": "model-00036-of-00042.safetensors", "model.layers.64.mlp.up_proj.weight": "model-00036-of-00042.safetensors", "model.layers.64.post_attention_layernorm.weight": "model-00036-of-00042.safetensors", "model.layers.64.self_attn.k_proj.weight": "model-00036-of-00042.safetensors", "model.layers.64.self_attn.o_proj.weight": "model-00036-of-00042.safetensors", "model.layers.64.self_attn.q_proj.weight": "model-00036-of-00042.safetensors", "model.layers.64.self_attn.v_proj.weight": "model-00036-of-00042.safetensors", "model.layers.65.input_layernorm.weight": "model-00036-of-00042.safetensors", "model.layers.65.mlp.down_proj.weight": "model-00036-of-00042.safetensors", "model.layers.65.mlp.gate_proj.weight": "model-00037-of-00042.safetensors", "model.layers.65.mlp.up_proj.weight": "model-00037-of-00042.safetensors", "model.layers.65.post_attention_layernorm.weight": "model-00037-of-00042.safetensors", "model.layers.65.self_attn.k_proj.weight": "model-00037-of-00042.safetensors", "model.layers.65.self_attn.o_proj.weight": "model-00037-of-00042.safetensors", "model.layers.65.self_attn.q_proj.weight": "model-00037-of-00042.safetensors", "model.layers.65.self_attn.v_proj.weight": "model-00037-of-00042.safetensors", "model.layers.66.input_layernorm.weight": "model-00037-of-00042.safetensors", "model.layers.66.mlp.down_proj.weight": "model-00037-of-00042.safetensors", "model.layers.66.mlp.gate_proj.weight": "model-00037-of-00042.safetensors", "model.layers.66.mlp.up_proj.weight": "model-00037-of-00042.safetensors", "model.layers.66.post_attention_layernorm.weight": "model-00037-of-00042.safetensors", "model.layers.66.self_attn.k_proj.weight": "model-00037-of-00042.safetensors", "model.layers.66.self_attn.o_proj.weight": "model-00037-of-00042.safetensors", "model.layers.66.self_attn.q_proj.weight": "model-00037-of-00042.safetensors", "model.layers.66.self_attn.v_proj.weight": "model-00037-of-00042.safetensors", "model.layers.67.input_layernorm.weight": "model-00037-of-00042.safetensors", "model.layers.67.mlp.down_proj.weight": "model-00038-of-00042.safetensors", "model.layers.67.mlp.gate_proj.weight": "model-00038-of-00042.safetensors", "model.layers.67.mlp.up_proj.weight": "model-00038-of-00042.safetensors", "model.layers.67.post_attention_layernorm.weight": "model-00038-of-00042.safetensors", "model.layers.67.self_attn.k_proj.weight": "model-00038-of-00042.safetensors", "model.layers.67.self_attn.o_proj.weight": "model-00038-of-00042.safetensors", "model.layers.67.self_attn.q_proj.weight": "model-00038-of-00042.safetensors", "model.layers.67.self_attn.v_proj.weight": "model-00038-of-00042.safetensors", "model.layers.68.input_layernorm.weight": "model-00038-of-00042.safetensors", "model.layers.68.mlp.down_proj.weight": "model-00038-of-00042.safetensors", "model.layers.68.mlp.gate_proj.weight": "model-00038-of-00042.safetensors", "model.layers.68.mlp.up_proj.weight": "model-00038-of-00042.safetensors", "model.layers.68.post_attention_layernorm.weight": "model-00038-of-00042.safetensors", "model.layers.68.self_attn.k_proj.weight": "model-00038-of-00042.safetensors", "model.layers.68.self_attn.o_proj.weight": "model-00039-of-00042.safetensors", "model.layers.68.self_attn.q_proj.weight": "model-00039-of-00042.safetensors", "model.layers.68.self_attn.v_proj.weight": "model-00039-of-00042.safetensors", "model.layers.69.input_layernorm.weight": "model-00039-of-00042.safetensors", "model.layers.69.mlp.down_proj.weight": "model-00039-of-00042.safetensors", "model.layers.69.mlp.gate_proj.weight": "model-00039-of-00042.safetensors", "model.layers.69.mlp.up_proj.weight": "model-00039-of-00042.safetensors", "model.layers.69.post_attention_layernorm.weight": "model-00039-of-00042.safetensors", "model.layers.69.self_attn.k_proj.weight": "model-00039-of-00042.safetensors", "model.layers.69.self_attn.o_proj.weight": "model-00039-of-00042.safetensors", "model.layers.69.self_attn.q_proj.weight": "model-00039-of-00042.safetensors", "model.layers.69.self_attn.v_proj.weight": "model-00039-of-00042.safetensors", "model.layers.7.input_layernorm.weight": "model-00039-of-00042.safetensors", "model.layers.7.mlp.down_proj.weight": "model-00039-of-00042.safetensors", "model.layers.7.mlp.gate_proj.weight": "model-00039-of-00042.safetensors", "model.layers.7.mlp.up_proj.weight": "model-00040-of-00042.safetensors", "model.layers.7.post_attention_layernorm.weight": "model-00040-of-00042.safetensors", "model.layers.7.self_attn.k_proj.weight": "model-00040-of-00042.safetensors", "model.layers.7.self_attn.o_proj.weight": "model-00040-of-00042.safetensors", "model.layers.7.self_attn.q_proj.weight": "model-00040-of-00042.safetensors", "model.layers.7.self_attn.v_proj.weight": "model-00040-of-00042.safetensors", "model.layers.70.input_layernorm.weight": "model-00040-of-00042.safetensors", "model.layers.70.mlp.down_proj.weight": "model-00040-of-00042.safetensors", "model.layers.70.mlp.gate_proj.weight": "model-00040-of-00042.safetensors", "model.layers.70.mlp.up_proj.weight": "model-00040-of-00042.safetensors", "model.layers.70.post_attention_layernorm.weight": "model-00040-of-00042.safetensors", "model.layers.70.self_attn.k_proj.weight": "model-00040-of-00042.safetensors", "model.layers.70.self_attn.o_proj.weight": "model-00040-of-00042.safetensors", "model.layers.70.self_attn.q_proj.weight": "model-00040-of-00042.safetensors", "model.layers.70.self_attn.v_proj.weight": "model-00040-of-00042.safetensors", "model.layers.71.input_layernorm.weight": "model-00040-of-00042.safetensors", "model.layers.71.mlp.down_proj.weight": "model-00040-of-00042.safetensors", "model.layers.71.mlp.gate_proj.weight": "model-00041-of-00042.safetensors", "model.layers.71.mlp.up_proj.weight": "model-00041-of-00042.safetensors", "model.layers.71.post_attention_layernorm.weight": "model-00041-of-00042.safetensors", "model.layers.71.self_attn.k_proj.weight": "model-00041-of-00042.safetensors", "model.layers.71.self_attn.o_proj.weight": "model-00041-of-00042.safetensors", "model.layers.71.self_attn.q_proj.weight": "model-00041-of-00042.safetensors", "model.layers.71.self_attn.v_proj.weight": "model-00041-of-00042.safetensors", "model.layers.8.input_layernorm.weight": "model-00041-of-00042.safetensors", "model.layers.8.mlp.down_proj.weight": "model-00041-of-00042.safetensors", "model.layers.8.mlp.gate_proj.weight": "model-00041-of-00042.safetensors", "model.layers.8.mlp.up_proj.weight": "model-00041-of-00042.safetensors", "model.layers.8.post_attention_layernorm.weight": "model-00041-of-00042.safetensors", "model.layers.8.self_attn.k_proj.weight": "model-00041-of-00042.safetensors", "model.layers.8.self_attn.o_proj.weight": "model-00041-of-00042.safetensors", "model.layers.8.self_attn.q_proj.weight": "model-00041-of-00042.safetensors", "model.layers.8.self_attn.v_proj.weight": "model-00041-of-00042.safetensors", "model.layers.9.input_layernorm.weight": "model-00041-of-00042.safetensors", "model.layers.9.mlp.down_proj.weight": "model-00042-of-00042.safetensors", "model.layers.9.mlp.gate_proj.weight": "model-00042-of-00042.safetensors", "model.layers.9.mlp.up_proj.weight": "model-00042-of-00042.safetensors", "model.layers.9.post_attention_layernorm.weight": "model-00042-of-00042.safetensors", "model.layers.9.self_attn.k_proj.weight": "model-00042-of-00042.safetensors", "model.layers.9.self_attn.o_proj.weight": "model-00042-of-00042.safetensors", "model.layers.9.self_attn.q_proj.weight": "model-00042-of-00042.safetensors", "model.layers.9.self_attn.v_proj.weight": "model-00042-of-00042.safetensors", "model.norm.weight": "model-00042-of-00042.safetensors"}}

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59f95e28944c062244741268596badc900df86c7f5ded05088d2da22a7379e06
+size 587583