Upload folder using huggingface_hub (#2)

- 798409a979d450e735696ebd1040081a798abba827e809137afcff8eab7acdf0 (c8a22b9cbec6d1b747bb9fd0122020dfb23e1026)
- a955320a8068680043c032ebd495ff5d90ad2382fdb0f696713f7c4ec67f7f13 (45aa27e1bb59424732295b32327c214cb0e35c21)
- e90bf7c7d894903acd370d10cad60d0ac50fc35681ac3f68a7ce313df9750cb8 (a4bc6bb5d32c965710538b46a475f69b978c1da8)
- d6ef2ac8dfca7a720e8c8813523ed7d85e6e49dbf59a4350bbd500d27000cd34 (c08340c35d6408e0693321ec5ee6d1e2bc9a1737)
- b692fe35e974ccf4e6471e5580c339aaee83d89258eadc43aa1a378ef80c6269 (60292a513778c38eb517274bf45f231809cc3397)
- 5b210dec1ad61e51869debb7cd22a8c984fa074a555b8bba40c4217ef4835884 (c3220398f104cb1e1b1c0584ff74f5cbaad05377)
- 2a075e57de335431ef072747491fdc9127cbb7d3125eac0f36e2819ed1928fce (0acacded49271a0d208977aad2b93be866665392)
- a6993eea56e56763af175dacd2e2fcb1bc18fe7da66ab8706c7b7e23fb9153a5 (f377db6e924e0e461ecb81e130017f1a13a3c47e)
- 77bd4dcb1a5a7f3b466c3a6d60b5e1530b69c0d2389d48f1214e5fffa8d7effd (7956978b2615b07de114adf5216d7ef1598f9cc5)
- 71cb2d9b4bd69f4a29d9bc1182ebb2fe55d991892558cb9610f341a4efe58bb8 (d81f456eaa4e7278b664e25b4d5f2ed105dba689)
- 5dc20712cc735d5029bfae52b76bb200bab3ee792b4141f51a4aaa653bb6e9f2 (8a2d7cab1cbaf899bac69f30f3e38e5b622417b7)
- 38e0013b1f4d29ad620e05c4fc8e7c7bdf5f9a3e493c838c9ada5cfc4716de5c (8bc22264786fb2fe57ae626db09662d0b5551840)
- 96ed557f923f5a5bdd69927cbd5cef5a2aadd66fe0e75f9b15af7df965f72e68 (65fb6320e37bec4dc68f7529d9bc4aab50a349af)
- 8a0d825c6450d033726a6752910d621e00b0292a743913b49c9f1351ec72d8e6 (974a5ceeb7ff0e383e98460fab2a6f3b8bfe37be)
- 0f886cee80c803d6bdadc90520de2a8588409ba731dde7677254ed96066a2f8c (69ee36cfa711a1ae24f7397b4659beaefa700843)
- 614bcedb4017ed5fe9f68fb8267a3d27f4e1a73ae99af3332d730897be05adac (2bbc2841b12b9ddc3e5aefce3feb7b4772380abc)
- d1d9de547b99d915b2552725477f0bdd48ea26a7cb2800b4cc4c87db46b859ae (9b68024941656baf574c653141328cef754a8a87)
- 1fe85f3368d26e48ed28e1e4a57392415441a36dc323c059b133264e655cd858 (9d56c8dd0dfc8d0922dd291b35a43ba49d31517c)
- 61c0ced816f09886269c46486ba64cc9b3dca88c9951a6f37028439d4450f125 (c5d986e94c051eaae5139b8bf1b80201236d0eff)
- cd7bf3aa23230bf276a5b5f855961f986c06be3ada1ed5bfb9b7bc4908686b85 (ffd457ffd507a112523c3a2d4e61441623ecd24d)
- efec8ef9ffc86848cac1f2e223c16dfd0e8e716cff88121bb80ec987cd4c76a8 (89aaf25bd62ce16cdeeba9236c37b40c11e91d81)
- 413ba75510ccd610ca06d2e8f97d56a06ab9c35e30eb0c7b10b94fe5e3c86f9c (d39024c630cb4734da6d7b7e93a80a08fa4c68f5)
- 44a0ff047671aa22ff068855caae617ec367e2a54b08607eeda7f985fb685abc (8f434a406efb850656c5c6014c92f1a52a432645)
- 69adbb8b0a0ad47065955247bed524ad611adba001a8633b5e342ad45ea5a0a9 (ba1f4706369c2b04bc319116a9f1fb4542c50661)
- a99a746a3f91797e99f57dabc8c7029af3a03192ecf88068f9899649a03256e9 (095e56a0e68b7a722f7d460eefe4b726c7e79754)
- 3b72a65957af1602a31d490747ec1157a33e7c8689b929697abab86e96a3ccc0 (a57c73ffbcc2a78e32c26ffbad00b7aad0f1e17d)
- 7e1675d0d1928f1adc92c7a9990772a67eeee8cd6f52aabefa04817472752749 (3b278bb60b10b0ae41f9e9abb004d97bcd3f4668)
- 844ad3692c9972783db89ec4d184052a199e05fda9a2b4120b2ec5ca48a07f91 (52c08f28d339517225b8a33802b0c3354cbc1dd3)
- 24abebb192240eb92013b9496908b17e991cd68178b36a74b1e51b22f8c738c1 (7341048a9df5316aa793700e59ec69ce7d6057ca)
- 2b23af56914d2fe770e9a61b08e8f865fd72d79a583137af704678de289c27dd (798352a59b28c59d8ea84a23507596139180b40e)
- a505d4506580310dafd838f835f957d858bcdc54bedd81efdfb71c61db1393ad (64acf850704cc81f73bac6ea0796c817642a33e8)
- fcdcb8b521623d44fa163263e407737877eea8f93071b4ae3f3abea9f4a27b2f (e03fb4a9052b75ffb23bedc65fde08f0ea0a5890)
- 4f03389a547f8dc3e87cc27d46c2b64c3c33c05b1f86b6e40e05eef512df6889 (8086d8f8229ed80a410c75159986f6912d0c338c)
- 78d3e1734d2b87d80b02f5fc717a002a4a731f98300c48437b3d3cba1b4291e8 (fc416c1bf419aff2a48f338531bdbbd30a7e94f0)
- f91824db826c51be7b4c574d1db256146f05301311ce603ac3371617d247992f (53ac90a76edf6c24921152b50f266d3c07535ebb)
- a6d370b598634e4ba607bd9e09e7ead6d4f1af6d845fd762c39f0bc15c22232f (c89ed9271cb8783ab90c694aaabaa7f09fa44c7d)
- 517420c08776b3ad19cee8392380da7d7f39dc39ad68f9eb8260b23bad343297 (42588e0bd7b98b08744bb74c4f9d3115beb916a4)
- 520f9ca27e83e87971ef989cbafe7b66f31e1afa199bbbc27ada47ace2c578b1 (2938cddc8ff5a8e53ea7c7b6e68fd35be6ca0f33)
- 6a42a43211582fdadf2936aacd2b39a86cc4a526918634b22db9d89094fb68da (a7897df04b5db2db1f4e75fae3370d00be42ab61)
- 8e16e42f487195dd7bfb05a8362e81d797057b43a2d94555e6df1791ac24e748 (730f8d57a87daee09147906303d4206f3dd85f98)
- 64199c1d077dafe8e662fa604f534ddee29a144b8022890d6e55bc90876132a1 (53ac39e023d4c83b8110401a042c6a4a8dbe8ab1)
- c948484a736333877f297355ea483adcd02a169834b722e7a7c42a89bd456780 (643b293f954c6342afc0f38fe47e84dbb3f224d8)
- 732f835f867672a13d878374c5eb07b1e14948cfe6513e9d43c4f220edc09348 (1ed127b3756bf594d390da3b3bccedc4f949b3b5)
- 27e9640da75a16d20de9c5681d466dcf0503b29da0abaa39ca3694a4e7bb37b7 (ffefcd3c9e632809470fe681778f804202967a95)
- d91b0db8fd4a95432063612843bf06eb0122961ba8998a6d8e378ba7d76e855e (6649c0947554b04debf8edca17d7f60e1a37ae1d)
- 1ccb648f918baaff7d4d982bbc2c779a0b8f33e185e8ca4d308ca43b5759af12 (d71ce08338876bde4285dd8952d4368d58b3c4f2)
- 0a489f67b4223c15cc5457d1ee4c2b25582594f34cf6885976f8b4f4925704bb (aef9400a825d37f8f79cf03aa8f6535d7449a1bf)
- f2fdde45bfa0889ddb462c2edce3b60943648c59fcbcca973e5b38b852c47aae (c268c54a2a7da08608ff297870a01ef0bf22d3bd)
- 8bbb28d9d92a358a76edc976620fe7cb51376d0f8f214d1273179607640dbd66 (3a6fbc1b3bc39401996e4424d72aec0a4574c680)
- 6f2f988a5ebd1e6c7f8dd5ecfbd6c3bf080d6812d89e388934e5414c6d17f113 (fb0b32ed330f823ab27610802df2a18763af7828)
- 8d70a6111eecc6753e1f42349356f9fdf7568bd41113af1c488c0ec10dbde705 (b54cb9291476344405d23c7ba776b55392c88027)
- 745ebf9e3a4ad88c995612acad31841bbe45ea78302b20dfdc9f8bd05e8156c3 (db8303fabe55c7e7d79fdfc24569aed9e70c68cd)
- 666fd69eedfc3e715a0ed5b230eba196e780aca68097076c5af4208f314ea6a7 (8c228bf9655ccd9946d93092d5c661a1c852952c)
- da77f193e87b2399af6b4ef25f520c0614c4943dbd99f95963d076da8f04c243 (1cf60d71713a8e9d296daa3b2620dca73aada0f3)
- 5199a10694939a782eb133e141f1e11b722e054f2498e6df65442cd7fa76a254 (0502516245a6858bc9f7e66243d21143df4a514b)
- 135faca87a89b08d7d0bd96ad741174cf758c7f15c945a546fa92d38735ecab5 (c22c55fae2dfd89516e36316f4514ea3e1c83567)
- c9ca4cb121c975a151698c242c86e47132c3b873e44b1078d20f57455c09f55e (6da1440fe8be2537aecdc4659be443ba9dd56e39)
- 9394a8e5522145824c0c86fbde9e39e2b1e42e1243b223e3a8f200640e0cfa2c (c9f8949246f21301590a5cfdeefa9ac178d2b803)
- 7de5d419872858434761d7f2d51d46e4c363d5da1ba8225c70d4821a1068cb27 (9e29fea84b3a1bee03dd6e27f7ae43eb73a8fd4e)
- e16c3245911b7c36da730eaac65105fdbe63e47b423283687ab4ff638992981a (84af724b3371b5db438f0094251a4e3e15fc457a)
- f95f93fe6e938a8c213401060925799137b9fcfd2deb84432065b04918407324 (5a8d5477a3ea89b3d06d7de0d898a23579aac42f)

.ipynb_checkpoints/config-checkpoint.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "sequelbox/StellarBright",
+  "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
+  }
+}

config.json

@@ -1,9 +1,10 @@
 {
-  "_name_or_path": "./new_model",
+  "_name_or_path": "sequelbox/StellarBright",
   "architectures": [
     "LlamaForCausalLM"
   ],
   "attention_bias": false,
+  "attention_dropout": 0.0,
   "bos_token_id": 1,
   "eos_token_id": 2,
   "hidden_act": "silu",
@@ -22,7 +23,7 @@
   "rope_theta": 10000.0,
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
-  "transformers_version": "4.34.0",
+  "transformers_version": "4.36.2",
   "use_cache": false,
   "vocab_size": 32000
 }

generation_config.json

@@ -3,6 +3,6 @@
   "bos_token_id": 1,
   "eos_token_id": 2,
   "pad_token_id": 0,
-  "transformers_version": "4.34.0",
+  "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:d2ac53bb4dae6c10c6944076c7f3ec095e3dec9fbdca3644a7f551153753b4e1
+size 4806739440

model-00002-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:121bcb109936f87f3b261d42626118b9c75a5232bb6001346b5f3c6b7715bb8b
+size 4630578440

model-00003-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42a045876286567cabad9e2939cadfd9a0567d25932ab5dc763bb737461d2040
+size 4362142864

model-00004-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab12cff6359e23672dc724a46cdba6ffa5779d7910571671eed5711b220d8165
+size 4966188864

model-00005-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c4e4bc72a7cd5d4b43bd8beb9b185779ce1decbe72d894fb70637768dd154e8
+size 4362142864

model-00006-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d3e461ca5e296ba89832180aac4e21a3beb6cf38eab8c49cfb0c2f7c90252d0
+size 4362142864

model-00007-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fd871e42bad877e03245ca91b1225a1dbb34b82c8a396f62e132498926ab43a
+size 4966188864

model-00008-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbb0e22bf6f924f1f85bb96418510f6c281784ba25a5e98ece5a97cface14fc4
+size 4362142880

model-00009-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7c29ede7d9edc1f3885e6b19100e5b9a9847b06c5dea58356d625ac45caa8da
+size 4362142872

model-00010-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:359f6cb3616691a1b948a38061abf8e427bda19582a96b67de4783a0b6e6f029
+size 4966188880

model-00011-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da3872d7cd89ccfb3cc928d08cc44802b6731fe743cecaeb500861d609a2efc4
+size 4362142872

model-00012-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99117139a8acf648ee2a15dc5c963c720bc952b6009e3452cdf0e25cb44df6da
+size 4362142872

model-00013-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfffb8c3b589ec25371adf87db888b260631ebddd2b872dfe49fc77a47811bd2
+size 4966188880

model-00014-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7fb55f5b1669e873730eda9af7f34befcf6a209b6b52aac6c66aee4747fcfb6
+size 4362142872

model-00015-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bcb7af4d98b855b238b9b75e70527dfa78e7625c420784ea1c9fceb3ae692d9
+size 4362142872

model-00016-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a95cf6d1372da205ee2ab8aae4cb1855125dbdd40126a2dfb556a2a39df3d977
+size 4966188880

model-00017-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b35ef866852ab8ab13fc456b12cb4bbff89144fb00f0c188e8006d2089ae3572
+size 4362142872

model-00018-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a52321fa890f893901ab5895c78f074ad3acbe1eaebe9f405e69cce1fa969d2
+size 4362142872

model-00019-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5d2d0e06ea60f75aefcc0263824aaaa8bb75a674cd371744342299bd1dc9179
+size 4966188880

model-00020-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e05162d9c2bbfc46576c5a980bfcd2dcbe891ed7aca953ccfeedfaf5d99979f9
+size 4362142872

model-00021-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15b69b49ef5bcbcc962b8a2909edb0ce7e47db53396823ecf87844c1b14f10af
+size 4362142872

model-00022-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e39c3030e24e1b81ba88661916b52a6d6e7cac08dbeb0fa1f7cfbe19cef30771
+size 4966188880

model-00023-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b009d7ac90308841277d8bdde931bcfe05ad9d30da73ac9b1690e432016f1b4d
+size 4362142872

model-00024-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acdfb1f2251422fbb1a36810f761842dc1290aeed0b00c9eb30fb182d15d9bff
+size 4362142872

model-00025-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26ced8103331d342cfd426249a9df401393e4c2a8db37129695b51425425dcc7
+size 4966188880

model-00026-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb5214ce3e36966dd1e49d392a94e654bcec397b3b1db8cb3b77b9ba6b1e4bbd
+size 4362142872

model-00027-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33829eb6e9ba3cdbee802de513ed092e9f347da746d365ab365e7527171e11a5
+size 4362142872

model-00028-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45021bda2a66f5cb44c7df685fc023bf0e6170b6a9f5281a276192592076cbe2
+size 4966188880

model-00029-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c151e105789827f7a9eb2d53cac848a8302736beb19db35ce245bda4fb572c8
+size 4362142872

model-00030-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3e936af2482f785f3597814ff4aa1adf3dbe302c5e6c5d59a1b3eb08ff90b93
+size 4362142872

model-00031-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19675dad772c48218244e91bf4070869644b36e0a381362575b45e6e070d7523
+size 4966188880

model-00032-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1900b2fe7d513a9b0272759aacce1bbb5c9f8cd4506078f0672fba4eb7da8401
+size 4362142872

model-00033-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e37dc160cc89083d93350742e74c731b8143d7e5c1932bc579aa1af398f244af
+size 4362142872

model-00034-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38c2c6068bb09d76d8e0ef52bd4ff4b8d514ceb0cec9dfa12a324adbdfa83206
+size 4966188880

model-00035-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50b7f73e0aa26595c54320efcc6eed0a5742f4d5910693f8f52a2fdcb0068e2f
+size 4362142872

model-00036-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d65101353e3fade85c0afe0a10408646ad2ee6be9f5c692c830a77555e627caf
+size 4362142872

model-00037-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef5d58b0b6aa95ffc46544c4a90ff50240b54833eda835c4902d1faad329182f
+size 4966188880

model-00038-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66e721362208d1c3f62cad3514d27d0e67631947194edf3642ca2bad7a4f7bbe
+size 4362142872

model-00039-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c3444c52f516a686360e63d5d477ea4faf743bc462d96c34dc8965557bf5304
+size 4362142872

model-00040-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8474c5811df68a0cd3d78094091a83266f260ae8b70abef3a31caa802cbf1543
+size 4966188880

model-00041-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1e9ba6eb12c60fb9bcce3ac299a163bac8c07ab3cfede4e52c55c8766f94b5b
+size 4362142872

model-00042-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f47596c5a91d8531d561a267f334dbc60f7b6ad8c5d221c82e18ad2dc4269a09
+size 4362142872

model-00043-of-00061.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41db8b4a083246ea066c06e0c6ae51c0208b098af581cb34781b9f547843e928
+size 4966188880