Upload folder using huggingface_hub (#1)

- 94af3249ee7d2f6b8be2c389a2a532e25abd3e942df5aa49194f06624d6ec7df (df15c8e7968db4a15de24735690f4a788b59ee08)
- 62ec0adb2e6eb61b84bf5753b8d55937675316bafc0e2fdf25d11d6795dfa25e (82f99016ca11306c72cf7870c5b47973a7025f84)
- 412d6cdaa4d4d74e8381bb114a41f7c6801177a2eb8de0cedd99f6dd1a6dbb6d (6c7dab0ec7726f28301dc8bbfa560ff13fdf7004)
- 848585d11936c4ae112bf93bdb59e5c759aa2a948a29f5b348a21692483ded0d (31c80b91151cec920f1e1e84442511cc0471a8be)
- b6261dfaf7c3a8aeabf16b4204d3fe0ab6a2973320e9698aa480f4c4fefc0bc8 (dc4a1e2d9e3026ae69a7331847617986f2b06484)
- b1190a762b308fca274cd5b94f07c056b291b57eb45e64c23c9f3eb0cb2eccb2 (e0cd4ed8bb9b1de55367bacd186b440055e3bc9c)
- 5c13dc3596290b66b847eb68af8eee320d5cdbedecd932ced9db3676f12116dd (7f5e4c9e784b130b829c0c3b06266221e6b80172)
- 63dc6b21fa0cb02b631da79d5c11168ccc310d4a7d03a71af390090340092d3c (4b295ca4ce303037c72b736676ac2627ebc91595)
- ebfd399034a4ddc90b959eb358013d1247f31ec237ee8b63fdf29bfaa0d440b5 (01ad1d15d43903ff0abd112ae43cec38e397663a)
- eb0a6533b36e84c8b2e7a36ebf5e5923b543c2a7653d628976333cbdc9e80aa0 (a6a3a3553a3406433b1adaa92c696fc41d3bec15)
- c20665c5f23317785cec4c9d0f7da76320d5236daba6c1a221334726d790eadd (1794d11175fcd29d7d50f1a897b64cfd9fab0137)
- ce858e8dc9990be378e9eb741afd4893a9a815a0e7f001cf9fbbc7a7a78984e9 (d1c1bd8260cb043c6c7ac7c6a295f49f3a42d7f9)
- ba3c47abdaaf7b18127dbedabac08e41a5615f84313f6e51150a6eb38ab4ec92 (1d4444e8c62513a7db86aa98c58d43dd86f2c250)
- 2a1440f1980ef647bc431f7d01a7fecd4311d733f161bad6670c4c06b3f93b02 (fbf5e925072bb59236fad297b687cfaad63bebec)
- 121a0045e764e701c2a4bb1b67be01fe422c5935e71f312e7cffdddf8c3ce91e (caebecd7f93cd533d7f519031cd0214d93e5d428)
- 00684905142fd5ca6093cae55cde0a0bf816a296294347cc63d59edfb648aed1 (f7e9ab61d55afdadd4ad9138a9f0f50da21c94bc)
- f2fca6496e27649572bb3d50976660aa6e1212bb29517e8bae7ea05766b57ad8 (40c9eeb13d757524bc0ba8e264ec44009e3a6ea0)
- 67271d62879cecdfd5ebe7180421781a589d111a944ae6915c018e1f2400884d (7c7cd95f70aae7657a1b8de6fc39ac09b1cce869)
- 3997b6ff6c69cb8c4fb2e8595c5d9aeb4cb813a9c26c482ad479d016f7d7e89b (ce4de135747ddf5e32415b6dc224d765b4e5ee2e)
- 09e08699082b878ed8d9a54acf2762db31c24d7edfa266a20640c042056ce41c (def6b545391b37f6e7d1eca6fa9a4d09f6114805)
- 9c2d4c0e364dbca0fff46e6fcc881682d7de168f3918563849a7fcf712d63ec0 (0a18b9779a44f50c0f9ed00cc4cb7e9dd5fd39de)
- 901c3c78b7f74b4022cd2745c381b08924dee6d5bb487b4eca4fe6ba382e8208 (7acd1f7688a10057cc3e2dcb68e710598a51d073)
- 8004e80664ae55918df6a582838a375c00a08fcd2d02a1d126f39c3a0742a36d (67aa82f8bfc304e58a2c17cdcf94970662d9567f)
- f8949fda5a613104702af6da5971c96c7fb3edd8c60d3eace9a6d55643480cce (3a03090755c5346ee85df97364d3a1bd736083b7)
- ebb0eb0f116d8704168107e0dc6b4448e1345782dee8d61d0a64cc46b0799466 (9fbe051ffbdc5e7bbd215575df943022018de970)
- 8785f45fc040fe79bd5cf21743c792e45d78f3197270e06da2d8c45e20533413 (698c045206713048d62b2c4d718b6a6ba99756ef)
- 741bdd74d3f2a553d717a94b1f0cd884ea959819cc455a33747ec14224ceace8 (1e9ced62984a72df54e2ad4f713e0a18c8b5ed68)
- ae44b464198fb42818814052f3f3b3881a7b33bbc5ebe633c1a71df83401fe38 (75d8e2058fcdd70fa913b1b28086e2fb9e8f86d4)
- 179c55eaa9f2e961752ab2fa32c63faa05f1e161396368566e663d641e9b2e95 (8345d3c8ab52c5fd8f25c4d1e9fd7f8baaddfc60)
- e2957fbd388fac7a32132f0e1b55975cbb769ba39f85a3e0db65af930604a609 (832219a75f6e30933a429bd59d0ddd7f2ffcd7d8)
- 6705812196fc694d4e1e9cf438d0e8e67ead54ecf0f0af5bdf5bf8e8f1e0eea1 (4d5629fe2ab00c24555feaa8793388d29463a265)
- 52837442612124c9c9bb5233b20375f107744a77620856b6fc0f1a60cc2b8c85 (7e871ffdb471d4cf98676d3ba288245692e12464)
- 13cafd840b0f1f35d94e8f743b29819b9b6763093851cd0a46a91aa1432e23be (a33e51a91c9f009a9d8fcb6f7bc107b202b67995)
- 3350024eeabd99f4dccc2c3824fc11da46ef3284199de7247fd1a079621696d7 (714c0a61e55148e9abeb0c26b00bb790e73ad78c)
- c9c9a18ead97df737c71124e5fcd02ac6e3360d0cf917c8a87e328f9c9fa1cc9 (e101b0649cd2ca454121c5c350aa85f04be16287)
- a103c940c961368b32ee2b306570bea3c6f02f223e3057d7a5e1ffb85412c700 (2b39a6facf0c93c37aa27b6a69ba31849b415f65)
- 372638ea9803605831171cac9add101fa21e78eb828b5d6c0cb5f794863e02e6 (b5436a2e078fe263ecef51ebc6dd8df2c21e557a)
- 039241a347695a6cbe01646105f544b8fafdf2c6b57a948bebe02b3b242e70c7 (b5ab169dbb361605c2956d14c90d38dfe16a24eb)
- d3464a3dda25bed02f33528d17eefa7a6174e2f310a42ff51cf081e59e18d00d (442f59e7e8208c650d6bcb2fb8a5e1b52d753f6b)
- 5860194d83aa6c31789555105659e6641529d8bf2ab998da9454a910a9ae2a2a (9f5ff064f5c0a48845b5b50b334202c81c1f39d1)
- fb8112f30fe67d8ed6127d87a546eaa5e2f4fcb41b1e60b0b615aaa13b75df20 (3f907711104336280aee10d3c92b400113b24115)
- 7399f9197a0ed39a9ae6fdffea87ef092d309c6f27dc096895929f20a32ab94f (c82f83089476a1ebef4bea363ed74562ae744140)
- ff1cf3a250ca5542cbeba0177fa3ff0c8e3cd8b106bce020d561bcc50c5255e7 (ebbb508ed49694f5a73451bd1c302e6005002777)
- b7340f108dfd263e1457bdf8cbf0747d97bf50e221a1b00ad9f6e066d1516ca3 (fee1e0f7582e1058d977d3c13f66a4ac5cfc5aaf)
- c64d4878e5b08398e3220973cfbcf653d51c7723c7945afaa5ac3189a7f866b3 (2335ab2c7a187d1bcc66df3ea354d098c0a2db10)
- fbe73ebc9846c7f284815109a4d5078243d7317feeaf9b18e4c1f892f565117f (0bd814d8a989082a4ecd8b451af509941d1c840b)
- cd5a0c152656f885ef2e0b70b9ef200ab1c1472f4d928e94b7a3098ac804b6bc (7b217ed8ea096c79bc41b61acb5938a259f38cad)
- a8cadbf0fe87dd61e102196e337da7b1a6e25ee009025ac6b159a583972cb55a (97c86e9dbc0be8598fbfa21a00e426529eb3df49)
- ac825f5998dec81cb71466d4bbc5c03cc83db88a27a9e870998928d45c245e58 (8cdcb62b0b3cfeb7abdd0d1e331c220deecd1dea)
- d5a8550526c35940bfcef1f0fb3af4065b96b847115e02c7d268ace2c98c0af3 (a73576e9b03235b4c4c657e6484eb5a68dfd7e93)
- fba5c2437f8af7d91bf3c30af0e9776747d0888ce69dcd3edd81c1097e7474e0 (7a2a960c2f1411cf506bf9181d1e900b48273e04)
- 90bb393b644f27c80abba53566ee0c8325ade8e86a08efc8b1f815a4fe6e8cdd (8948e27f1942cb9237360feb40f37061c2ad49a9)
- 3648ab95925cf7ea5ecba4525589ba464f4c321beca1f47b15154f075ba87fa1 (101e11c5a860fc7e8346c4da5f7bda4088ffbb72)
- fe904876f884f632ddcd034acec523bfb870b804775c3ed1a40389cd9dc67a7d (e8aae75634eb0b35aa811ecfad83cd123bec3c0e)
- 0024f1ce915d6e407473592476b05038f8cdb0709d2a4aa331c9f30a00d9e487 (92efb423c37759fd50aa0aea1db9e4c12f3aac98)
- 5998b27b46112b6ce6f2df515c9a1b9d212e3e75ffb7c35966680964b85a449d (a4a5d884f06a5cecee1395ff7a2d89d3283d646c)
- 61e15c2eff9b3564b8f076bcb650b9e9847d7aabc386eda7d77f10e8c49f1136 (856f0b2b9826d2b51bf171ea647d92e90fca715f)
- 3a2f406c7c2cc7de0cf4a198623824c979809a74ec4f26dad2aa47189b42215f (bb381012b149bd75e965712581ce299c0f544612)
- de3669e9b655c76dbc22feeee91d4caea1fbf028c9aacb565d435fdf2319c5d5 (8005c4cc90add8c3811683b2a55f0bb411165bc4)
- f0868b81529d08b281ef5b27f1d312282532a034d546729fd5d11129b47a1a51 (d4ba2af71396c2e88dcdcc7591e89b7b197d0d0d)
- 76d9822d3fe37ddc05aeede0658bd020de5c5917050cde9098ec4db175c12956 (1f64b191bbbca53912f938ed10b67e866c54a59d)

.ipynb_checkpoints/config-checkpoint.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "sequelbox/SpellBlade",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 8192,
+  "initializer_range": 0.02,
+  "intermediate_size": 28672,
+  "max_position_embeddings": 4096,
+  "model_type": "llama",
+  "num_attention_heads": 64,
+  "num_hidden_layers": 80,
+  "num_key_value_heads": 8,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "use_cache": false,
+  "vocab_size": 32000
+}

.ipynb_checkpoints/generation_config-checkpoint.json

@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.36.2",
+  "use_cache": false
+}

.ipynb_checkpoints/model.safetensors.index-checkpoint.json

@@ -0,0 +1,730 @@
+{
+  "metadata": {
+    "total_size": 275906592768
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00061-of-00061.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00002-of-00061.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00061.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00061.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00009-of-00061.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00009-of-00061.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00010-of-00061.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00010-of-00061.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00009-of-00061.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00011-of-00061.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00011-of-00061.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00010-of-00061.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00012-of-00061.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00012-of-00061.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00011-of-00061.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00013-of-00061.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00013-of-00061.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00012-of-00061.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00014-of-00061.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00014-of-00061.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00013-of-00061.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00015-of-00061.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00015-of-00061.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00014-of-00061.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00016-of-00061.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00016-of-00061.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00015-of-00061.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00003-of-00061.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00003-of-00061.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00061.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00017-of-00061.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00017-of-00061.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00016-of-00061.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00018-of-00061.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00018-of-00061.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00017-of-00061.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00019-of-00061.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00019-of-00061.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00018-of-00061.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00020-of-00061.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00020-of-00061.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00019-of-00061.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00021-of-00061.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00021-of-00061.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00020-of-00061.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00022-of-00061.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00022-of-00061.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00021-of-00061.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00023-of-00061.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00023-of-00061.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00022-of-00061.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00004-of-00061.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00004-of-00061.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00003-of-00061.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00024-of-00061.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00024-of-00061.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00023-of-00061.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00025-of-00061.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00025-of-00061.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00024-of-00061.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00026-of-00061.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00026-of-00061.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00025-of-00061.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00027-of-00061.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00027-of-00061.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00026-of-00061.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00028-of-00061.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00028-of-00061.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00027-of-00061.safetensors",
+    "model.layers.36.input_layernorm.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.mlp.down_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.mlp.gate_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.mlp.up_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.post_attention_layernorm.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.self_attn.k_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.self_attn.o_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.self_attn.q_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.36.self_attn.v_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.37.input_layernorm.weight": "model-00029-of-00061.safetensors",
+    "model.layers.37.mlp.down_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.37.mlp.gate_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.37.mlp.up_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.37.post_attention_layernorm.weight": "model-00029-of-00061.safetensors",
+    "model.layers.37.self_attn.k_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.37.self_attn.o_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.37.self_attn.q_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.37.self_attn.v_proj.weight": "model-00028-of-00061.safetensors",
+    "model.layers.38.input_layernorm.weight": "model-00030-of-00061.safetensors",
+    "model.layers.38.mlp.down_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.38.mlp.gate_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.38.mlp.up_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.38.post_attention_layernorm.weight": "model-00030-of-00061.safetensors",
+    "model.layers.38.self_attn.k_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.38.self_attn.o_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.38.self_attn.q_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.38.self_attn.v_proj.weight": "model-00029-of-00061.safetensors",
+    "model.layers.39.input_layernorm.weight": "model-00031-of-00061.safetensors",
+    "model.layers.39.mlp.down_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.39.mlp.gate_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.39.mlp.up_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.39.post_attention_layernorm.weight": "model-00031-of-00061.safetensors",
+    "model.layers.39.self_attn.k_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.39.self_attn.o_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.39.self_attn.q_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.39.self_attn.v_proj.weight": "model-00030-of-00061.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.40.input_layernorm.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.mlp.down_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.mlp.gate_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.mlp.up_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.post_attention_layernorm.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.self_attn.k_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.self_attn.o_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.self_attn.q_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.40.self_attn.v_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.41.input_layernorm.weight": "model-00032-of-00061.safetensors",
+    "model.layers.41.mlp.down_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.41.mlp.gate_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.41.mlp.up_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.41.post_attention_layernorm.weight": "model-00032-of-00061.safetensors",
+    "model.layers.41.self_attn.k_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.41.self_attn.o_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.41.self_attn.q_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.41.self_attn.v_proj.weight": "model-00031-of-00061.safetensors",
+    "model.layers.42.input_layernorm.weight": "model-00033-of-00061.safetensors",
+    "model.layers.42.mlp.down_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.42.mlp.gate_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.42.mlp.up_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.42.post_attention_layernorm.weight": "model-00033-of-00061.safetensors",
+    "model.layers.42.self_attn.k_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.42.self_attn.o_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.42.self_attn.q_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.42.self_attn.v_proj.weight": "model-00032-of-00061.safetensors",
+    "model.layers.43.input_layernorm.weight": "model-00034-of-00061.safetensors",
+    "model.layers.43.mlp.down_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.43.mlp.gate_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.43.mlp.up_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.43.post_attention_layernorm.weight": "model-00034-of-00061.safetensors",
+    "model.layers.43.self_attn.k_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.43.self_attn.o_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.43.self_attn.q_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.43.self_attn.v_proj.weight": "model-00033-of-00061.safetensors",
+    "model.layers.44.input_layernorm.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.mlp.down_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.mlp.gate_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.mlp.up_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.post_attention_layernorm.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.self_attn.k_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.self_attn.o_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.self_attn.q_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.44.self_attn.v_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.45.input_layernorm.weight": "model-00035-of-00061.safetensors",
+    "model.layers.45.mlp.down_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.45.mlp.gate_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.45.mlp.up_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.45.post_attention_layernorm.weight": "model-00035-of-00061.safetensors",
+    "model.layers.45.self_attn.k_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.45.self_attn.o_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.45.self_attn.q_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.45.self_attn.v_proj.weight": "model-00034-of-00061.safetensors",
+    "model.layers.46.input_layernorm.weight": "model-00036-of-00061.safetensors",
+    "model.layers.46.mlp.down_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.46.mlp.gate_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.46.mlp.up_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.46.post_attention_layernorm.weight": "model-00036-of-00061.safetensors",
+    "model.layers.46.self_attn.k_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.46.self_attn.o_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.46.self_attn.q_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.46.self_attn.v_proj.weight": "model-00035-of-00061.safetensors",
+    "model.layers.47.input_layernorm.weight": "model-00037-of-00061.safetensors",
+    "model.layers.47.mlp.down_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.47.mlp.gate_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.47.mlp.up_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.47.post_attention_layernorm.weight": "model-00037-of-00061.safetensors",
+    "model.layers.47.self_attn.k_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.47.self_attn.o_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.47.self_attn.q_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.47.self_attn.v_proj.weight": "model-00036-of-00061.safetensors",
+    "model.layers.48.input_layernorm.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.mlp.down_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.mlp.gate_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.mlp.up_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.post_attention_layernorm.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.self_attn.k_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.self_attn.o_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.self_attn.q_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.48.self_attn.v_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.49.input_layernorm.weight": "model-00038-of-00061.safetensors",
+    "model.layers.49.mlp.down_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.49.mlp.gate_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.49.mlp.up_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.49.post_attention_layernorm.weight": "model-00038-of-00061.safetensors",
+    "model.layers.49.self_attn.k_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.49.self_attn.o_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.49.self_attn.q_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.49.self_attn.v_proj.weight": "model-00037-of-00061.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00005-of-00061.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00005-of-00061.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00004-of-00061.safetensors",
+    "model.layers.50.input_layernorm.weight": "model-00039-of-00061.safetensors",
+    "model.layers.50.mlp.down_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.50.mlp.gate_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.50.mlp.up_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.50.post_attention_layernorm.weight": "model-00039-of-00061.safetensors",
+    "model.layers.50.self_attn.k_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.50.self_attn.o_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.50.self_attn.q_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.50.self_attn.v_proj.weight": "model-00038-of-00061.safetensors",
+    "model.layers.51.input_layernorm.weight": "model-00040-of-00061.safetensors",
+    "model.layers.51.mlp.down_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.51.mlp.gate_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.51.mlp.up_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.51.post_attention_layernorm.weight": "model-00040-of-00061.safetensors",
+    "model.layers.51.self_attn.k_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.51.self_attn.o_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.51.self_attn.q_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.51.self_attn.v_proj.weight": "model-00039-of-00061.safetensors",
+    "model.layers.52.input_layernorm.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.mlp.down_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.mlp.gate_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.mlp.up_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.post_attention_layernorm.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.self_attn.k_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.self_attn.o_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.self_attn.q_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.52.self_attn.v_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.53.input_layernorm.weight": "model-00041-of-00061.safetensors",
+    "model.layers.53.mlp.down_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.53.mlp.gate_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.53.mlp.up_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.53.post_attention_layernorm.weight": "model-00041-of-00061.safetensors",
+    "model.layers.53.self_attn.k_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.53.self_attn.o_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.53.self_attn.q_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.53.self_attn.v_proj.weight": "model-00040-of-00061.safetensors",
+    "model.layers.54.input_layernorm.weight": "model-00042-of-00061.safetensors",
+    "model.layers.54.mlp.down_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.54.mlp.gate_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.54.mlp.up_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.54.post_attention_layernorm.weight": "model-00042-of-00061.safetensors",
+    "model.layers.54.self_attn.k_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.54.self_attn.o_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.54.self_attn.q_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.54.self_attn.v_proj.weight": "model-00041-of-00061.safetensors",
+    "model.layers.55.input_layernorm.weight": "model-00043-of-00061.safetensors",
+    "model.layers.55.mlp.down_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.55.mlp.gate_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.55.mlp.up_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.55.post_attention_layernorm.weight": "model-00043-of-00061.safetensors",
+    "model.layers.55.self_attn.k_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.55.self_attn.o_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.55.self_attn.q_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.55.self_attn.v_proj.weight": "model-00042-of-00061.safetensors",
+    "model.layers.56.input_layernorm.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.mlp.down_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.mlp.gate_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.mlp.up_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.post_attention_layernorm.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.self_attn.k_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.self_attn.o_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.self_attn.q_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.56.self_attn.v_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.57.input_layernorm.weight": "model-00044-of-00061.safetensors",
+    "model.layers.57.mlp.down_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.57.mlp.gate_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.57.mlp.up_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.57.post_attention_layernorm.weight": "model-00044-of-00061.safetensors",
+    "model.layers.57.self_attn.k_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.57.self_attn.o_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.57.self_attn.q_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.57.self_attn.v_proj.weight": "model-00043-of-00061.safetensors",
+    "model.layers.58.input_layernorm.weight": "model-00045-of-00061.safetensors",
+    "model.layers.58.mlp.down_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.58.mlp.gate_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.58.mlp.up_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.58.post_attention_layernorm.weight": "model-00045-of-00061.safetensors",
+    "model.layers.58.self_attn.k_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.58.self_attn.o_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.58.self_attn.q_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.58.self_attn.v_proj.weight": "model-00044-of-00061.safetensors",
+    "model.layers.59.input_layernorm.weight": "model-00046-of-00061.safetensors",
+    "model.layers.59.mlp.down_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.59.mlp.gate_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.59.mlp.up_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.59.post_attention_layernorm.weight": "model-00046-of-00061.safetensors",
+    "model.layers.59.self_attn.k_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.59.self_attn.o_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.59.self_attn.q_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.59.self_attn.v_proj.weight": "model-00045-of-00061.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00006-of-00061.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00006-of-00061.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00005-of-00061.safetensors",
+    "model.layers.60.input_layernorm.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.mlp.down_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.mlp.gate_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.mlp.up_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.post_attention_layernorm.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.self_attn.k_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.self_attn.o_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.self_attn.q_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.60.self_attn.v_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.61.input_layernorm.weight": "model-00047-of-00061.safetensors",
+    "model.layers.61.mlp.down_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.61.mlp.gate_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.61.mlp.up_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.61.post_attention_layernorm.weight": "model-00047-of-00061.safetensors",
+    "model.layers.61.self_attn.k_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.61.self_attn.o_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.61.self_attn.q_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.61.self_attn.v_proj.weight": "model-00046-of-00061.safetensors",
+    "model.layers.62.input_layernorm.weight": "model-00048-of-00061.safetensors",
+    "model.layers.62.mlp.down_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.62.mlp.gate_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.62.mlp.up_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.62.post_attention_layernorm.weight": "model-00048-of-00061.safetensors",
+    "model.layers.62.self_attn.k_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.62.self_attn.o_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.62.self_attn.q_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.62.self_attn.v_proj.weight": "model-00047-of-00061.safetensors",
+    "model.layers.63.input_layernorm.weight": "model-00049-of-00061.safetensors",
+    "model.layers.63.mlp.down_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.63.mlp.gate_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.63.mlp.up_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.63.post_attention_layernorm.weight": "model-00049-of-00061.safetensors",
+    "model.layers.63.self_attn.k_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.63.self_attn.o_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.63.self_attn.q_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.63.self_attn.v_proj.weight": "model-00048-of-00061.safetensors",
+    "model.layers.64.input_layernorm.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.mlp.down_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.mlp.gate_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.mlp.up_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.post_attention_layernorm.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.self_attn.k_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.self_attn.o_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.self_attn.q_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.64.self_attn.v_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.65.input_layernorm.weight": "model-00050-of-00061.safetensors",
+    "model.layers.65.mlp.down_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.65.mlp.gate_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.65.mlp.up_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.65.post_attention_layernorm.weight": "model-00050-of-00061.safetensors",
+    "model.layers.65.self_attn.k_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.65.self_attn.o_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.65.self_attn.q_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.65.self_attn.v_proj.weight": "model-00049-of-00061.safetensors",
+    "model.layers.66.input_layernorm.weight": "model-00051-of-00061.safetensors",
+    "model.layers.66.mlp.down_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.66.mlp.gate_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.66.mlp.up_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.66.post_attention_layernorm.weight": "model-00051-of-00061.safetensors",
+    "model.layers.66.self_attn.k_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.66.self_attn.o_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.66.self_attn.q_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.66.self_attn.v_proj.weight": "model-00050-of-00061.safetensors",
+    "model.layers.67.input_layernorm.weight": "model-00052-of-00061.safetensors",
+    "model.layers.67.mlp.down_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.67.mlp.gate_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.67.mlp.up_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.67.post_attention_layernorm.weight": "model-00052-of-00061.safetensors",
+    "model.layers.67.self_attn.k_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.67.self_attn.o_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.67.self_attn.q_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.67.self_attn.v_proj.weight": "model-00051-of-00061.safetensors",
+    "model.layers.68.input_layernorm.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.mlp.down_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.mlp.gate_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.mlp.up_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.post_attention_layernorm.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.self_attn.k_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.self_attn.o_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.self_attn.q_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.68.self_attn.v_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.69.input_layernorm.weight": "model-00053-of-00061.safetensors",
+    "model.layers.69.mlp.down_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.69.mlp.gate_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.69.mlp.up_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.69.post_attention_layernorm.weight": "model-00053-of-00061.safetensors",
+    "model.layers.69.self_attn.k_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.69.self_attn.o_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.69.self_attn.q_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.69.self_attn.v_proj.weight": "model-00052-of-00061.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00007-of-00061.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00007-of-00061.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00006-of-00061.safetensors",
+    "model.layers.70.input_layernorm.weight": "model-00054-of-00061.safetensors",
+    "model.layers.70.mlp.down_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.70.mlp.gate_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.70.mlp.up_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.70.post_attention_layernorm.weight": "model-00054-of-00061.safetensors",
+    "model.layers.70.self_attn.k_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.70.self_attn.o_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.70.self_attn.q_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.70.self_attn.v_proj.weight": "model-00053-of-00061.safetensors",
+    "model.layers.71.input_layernorm.weight": "model-00055-of-00061.safetensors",
+    "model.layers.71.mlp.down_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.71.mlp.gate_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.71.mlp.up_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.71.post_attention_layernorm.weight": "model-00055-of-00061.safetensors",
+    "model.layers.71.self_attn.k_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.71.self_attn.o_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.71.self_attn.q_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.71.self_attn.v_proj.weight": "model-00054-of-00061.safetensors",
+    "model.layers.72.input_layernorm.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.mlp.down_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.mlp.gate_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.mlp.up_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.post_attention_layernorm.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.self_attn.k_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.self_attn.o_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.self_attn.q_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.72.self_attn.v_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.73.input_layernorm.weight": "model-00056-of-00061.safetensors",
+    "model.layers.73.mlp.down_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.73.mlp.gate_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.73.mlp.up_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.73.post_attention_layernorm.weight": "model-00056-of-00061.safetensors",
+    "model.layers.73.self_attn.k_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.73.self_attn.o_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.73.self_attn.q_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.73.self_attn.v_proj.weight": "model-00055-of-00061.safetensors",
+    "model.layers.74.input_layernorm.weight": "model-00057-of-00061.safetensors",
+    "model.layers.74.mlp.down_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.74.mlp.gate_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.74.mlp.up_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.74.post_attention_layernorm.weight": "model-00057-of-00061.safetensors",
+    "model.layers.74.self_attn.k_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.74.self_attn.o_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.74.self_attn.q_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.74.self_attn.v_proj.weight": "model-00056-of-00061.safetensors",
+    "model.layers.75.input_layernorm.weight": "model-00058-of-00061.safetensors",
+    "model.layers.75.mlp.down_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.75.mlp.gate_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.75.mlp.up_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.75.post_attention_layernorm.weight": "model-00058-of-00061.safetensors",
+    "model.layers.75.self_attn.k_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.75.self_attn.o_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.75.self_attn.q_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.75.self_attn.v_proj.weight": "model-00057-of-00061.safetensors",
+    "model.layers.76.input_layernorm.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.mlp.down_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.mlp.gate_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.mlp.up_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.post_attention_layernorm.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.self_attn.k_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.self_attn.o_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.self_attn.q_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.76.self_attn.v_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.77.input_layernorm.weight": "model-00059-of-00061.safetensors",
+    "model.layers.77.mlp.down_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.77.mlp.gate_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.77.mlp.up_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.77.post_attention_layernorm.weight": "model-00059-of-00061.safetensors",
+    "model.layers.77.self_attn.k_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.77.self_attn.o_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.77.self_attn.q_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.77.self_attn.v_proj.weight": "model-00058-of-00061.safetensors",
+    "model.layers.78.input_layernorm.weight": "model-00060-of-00061.safetensors",
+    "model.layers.78.mlp.down_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.78.mlp.gate_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.78.mlp.up_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.78.post_attention_layernorm.weight": "model-00060-of-00061.safetensors",
+    "model.layers.78.self_attn.k_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.78.self_attn.o_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.78.self_attn.q_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.78.self_attn.v_proj.weight": "model-00059-of-00061.safetensors",
+    "model.layers.79.input_layernorm.weight": "model-00061-of-00061.safetensors",
+    "model.layers.79.mlp.down_proj.weight": "model-00061-of-00061.safetensors",
+    "model.layers.79.mlp.gate_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.79.mlp.up_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.79.post_attention_layernorm.weight": "model-00061-of-00061.safetensors",
+    "model.layers.79.self_attn.k_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.79.self_attn.o_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.79.self_attn.q_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.79.self_attn.v_proj.weight": "model-00060-of-00061.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00008-of-00061.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00008-of-00061.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00008-of-00061.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00007-of-00061.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00007-of-00061.safetensors",
+    "model.norm.weight": "model-00061-of-00061.safetensors"
+  }
+}

.ipynb_checkpoints/special_tokens_map-checkpoint.json

@@ -0,0 +1,24 @@
+{
+  "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
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

.ipynb_checkpoints/tokenizer_config-checkpoint.json

@@ -0,0 +1,40 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": false,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": true
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "sequelbox/SpellBlade",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 8192,
+  "initializer_range": 0.02,
+  "intermediate_size": 28672,
+  "max_position_embeddings": 4096,
+  "model_type": "llama",
+  "num_attention_heads": 64,
+  "num_hidden_layers": 80,
+  "num_key_value_heads": 8,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "use_cache": false,
+  "vocab_size": 32000
+}

generation_config.json

@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.36.2",
+  "use_cache": false
+}

model-00001-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5352d1d26fb1a4187e6655674d3d7b9e65578dd4f88e05e3053008c53acfe878
+size 4806739440

model-00002-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e28499b8d3d1d6db61e3385c3736a1d8faaf891a2fc2ce89b71739f3b23a551
+size 4630578440

model-00003-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a69dced67f39f9f6c60816642904fba0e123102682df36a7703c4da3ed2b3478
+size 4362142864

model-00004-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c06846f87b3828821532d315409db90421ec463a8fb3057a2e28d5ebff4b368b
+size 4966188864

model-00005-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da16ef7dba13ec0c23aa5721c1c39a1793f0aac98a122942afb50bfd33ab0d12
+size 4362142864

model-00006-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7af94011077c5ff5cb6e41b28360c9ddecb491db47380274a61e687cea70001
+size 4362142864

model-00007-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54c6b0177452845eef4a53ff47787150687efe0e8ae05cf3bcdb45ae9b85ce95
+size 4966188864

model-00008-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b60807d26c02d8c09f84aae1e630e4c298eec45b90bc8a9bb4ff21b9886a65f0
+size 4362142880

model-00009-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20fafa021662573a48905ec08221dd85ed0181a0088055832fa95997f92cbed6
+size 4362142872

model-00010-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d86186d95a9895d1c2942dedaa390c45b8749fa7dc816461ee45fc08c764b4ec
+size 4966188880

model-00011-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70fa9d2b4d1922ed02f61b527528d56444692ec3fc136998eb06eeb344194016
+size 4362142872

model-00012-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e696bdf1acd81583bd846d5355bd401876398bc3b1ba8977846a1125dcaf1cc6
+size 4362142872

model-00013-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0eed999502c6eeff2b1a5158b16f7db3d3121c957ec9143eea3552eef8562082
+size 4966188880

model-00014-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ed50e80d456ce42bab05404bc5a13f36cd668a8acfd055cbe945647b79bdf69
+size 4362142872

model-00015-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24534c62918587285be9533a4cde711864570ecc425350fbb0278a591f2c6393
+size 4362142872

model-00016-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:417705ff101a6516ed3d3a018fe5968109e6505eaf88966233fc3485956a1efa
+size 4966188880

model-00017-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50ac743e5437c9321360795d533c3767a4454ba373d0ab781970c5b67d871568
+size 4362142872

model-00018-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35bb3eb4e4d049834b41635404827f9df27ab3c996aa9734a25da421c200c2a0
+size 4362142872

model-00019-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1cedc43136ff86f562bc67c9088d54e9b8e495c0052f31e975dda03d715a0c1
+size 4966188880

model-00020-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:692706f2b9e2c65accd5e8c2cafbb12447f9b97968c30ceb7645de7f70a5111f
+size 4362142872

model-00021-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70f933252a5599e0388aeeaf126f2a3c14f9be83724ed8a08bad491511ec7269
+size 4362142872

model-00022-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:454788a5f8ed3ca95d84b6fae013ab9832d517c1714239240e80016d7a1c4d15
+size 4966188880

model-00023-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4aefb059c46a9998dfc942961bfeedc719e854ed1189c0b4102709acf297e539
+size 4362142872

model-00024-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15de4c62eeee6d8c6c58039c17c9dc1d298b73e93838880c5d8242842b75b05f
+size 4362142872

model-00025-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f8a8d5a10febcf110b98976ffcec9ca8cd17d1fda8d7046b6726b04728b412c
+size 4966188880

model-00026-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4afd46ecebb697ade9f2a6d9323a635e43bbb1f3364c91fe3434162eaffdb8af
+size 4362142872

model-00027-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c7c91daeae73cfdbcb3167604a4e002fb3803e4ed0e21a5c9cb1fde61688835
+size 4362142872

model-00028-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:602add813be1febc8a4330796075ad18927aea3c5215dea0f976fd8a87cac61f
+size 4966188880

model-00029-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9d2db54aecdf9bdf854cb640be346b4023ef2857abeb720297c59627a7ea8e2
+size 4362142872

model-00030-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c49e0ba95fa93516bb09eb6d6d8cdc1c9c36d9f11ea24d64834e5a81817a41a9
+size 4362142872

model-00031-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74afde4aa694a3f0e7568bfc5c2cc72d14e632fb0b8c1c0b52cc34131588d03e
+size 4966188880

model-00032-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c553a79c4dc7527158d1ee8c63e6cfdfed7cb544247e14f8a262a869db928ac1
+size 4362142872

model-00033-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e64f4c3e6c63b879ab412b1b10c81411e2a885deea883c4d1cc1392e6ad06ee6
+size 4362142872

model-00034-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a7c8d6bf49315f4a15792aad78c45ad306f997fc870fa639e82c2411c361362
+size 4966188880

model-00035-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d57fb6faaecc3f1ce6646bd3c1576712e7e00e82bebc099475710f1a30ed133e
+size 4362142872

model-00036-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2543b012997bc2f260198166c51c328e98edd2e4a764317c43147aa6ee1834d6
+size 4362142872

model-00037-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2caf2871528d1a8a5871d90208b53312e3b821f09ca8c9aa6ac7ef8e55fc6a4
+size 4966188880

model-00038-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:818c322947e1675269b0e4513cc8cb8168849f46e9ce8c4bec12b57c6a5a3fbe
+size 4362142872

model-00039-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0f4aae1fc83d6c5a7235e7255b5f86c28b42b1077d1237878b59bcc0ac53b86
+size 4362142872

model-00040-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5ad4528388d48d8854250724ad0025e05091bb4af040186d0cd11aa508e7a90
+size 4966188880

model-00041-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8bf95940073d1e160e5349dbacafff6aa9b93d56ee7dc17500462c38b913dec
+size 4362142872

model-00042-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1f30feb65fd41e48016368c5c12a077ae2c2c01cb2690bf465288c9141eefaa
+size 4362142872