diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fde85b5e23419d2c8703ca7cf2fbc64b4d13eb24 --- /dev/null +++ b/config.json @@ -0,0 +1,40 @@ +{ + "_name_or_path": "/raid/checkpoint/averitec/Meta-Llama-3.1-70B-HerO-bs/checkpoint-48", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": [ + 128001, + 128008, + 128009 + ], + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 8192, + "initializer_range": 0.02, + "intermediate_size": 28672, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 64, + "num_hidden_layers": 80, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 8.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": false, + "torch_dtype": "float32", + "transformers_version": "4.46.1", + "use_cache": false, + "vocab_size": 128256 +} diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0484b997a9ea9b5b6d711db644716bfd32d5470e --- /dev/null +++ b/generation_config.json @@ -0,0 +1,12 @@ +{ + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": [ + 128001, + 128008, + 128009 + ], + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.46.1" +} diff --git a/model-00001-of-00062.safetensors b/model-00001-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..02109f5a30fb2e8681322a9f82f610b248ad1e79 --- /dev/null +++ b/model-00001-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5b4c66ef691095de5bef4270fe0613a4392c3dcae5190a82664c64772472adeb +size 4806672984 diff --git a/model-00002-of-00062.safetensors b/model-00002-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..dbead6521f90530af6185460fb78f380de89f1db --- /dev/null +++ b/model-00002-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e569256fe933a4d160c2231e251278ca0cea04ece1e21544a7f71f333db0c48a +size 4362142864 diff --git a/model-00003-of-00062.safetensors b/model-00003-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f223c0949c381fec4ca410150d81d63f891ca10c --- /dev/null +++ b/model-00003-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d8746592cbcb1d66f65490349cb9c21131fab3ce6aa20576723b1f25bbf3c7b +size 4362142864 diff --git a/model-00004-of-00062.safetensors b/model-00004-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9b7295e766396153ac67be72319c435e4833b2ef --- /dev/null +++ b/model-00004-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ef3a52c53bc55852195e964b2716b196ac789c99ad8fcf82b627690fb49494cf +size 4966188864 diff --git a/model-00005-of-00062.safetensors b/model-00005-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ffd83e2d9f0c37e9e44aedde7bd16245b8ad3f95 --- /dev/null +++ b/model-00005-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be0e80da994ea0de80729d420c6332a833b2d107ec92d0b26af0b206ae2694e9 +size 4362142864 diff --git a/model-00006-of-00062.safetensors b/model-00006-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..00d835f0fe60cf472847c131b1571ca7d13a6b76 --- /dev/null +++ b/model-00006-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f1afec345058951a18cd4b5ddc0654b1a5436bc572b046f056c795b62786300d +size 4362142864 diff --git a/model-00007-of-00062.safetensors b/model-00007-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..13a48c2769fc151710f7b3459d2ef95d90723eaa --- /dev/null +++ b/model-00007-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:14b9e367fa75d6f0746f23a853495b2a67429ecea58251463821b193e7acbae7 +size 4966188864 diff --git a/model-00008-of-00062.safetensors b/model-00008-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f247a8a9aa7fe1360c6fd4e37628262c89939197 --- /dev/null +++ b/model-00008-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91e996843e886ff412b9248027120b6443e13f56cabaa3ca550b201aeaa729b6 +size 4362142864 diff --git a/model-00009-of-00062.safetensors b/model-00009-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aff484592d66d73900ec485096db5b599ee4b1bb --- /dev/null +++ b/model-00009-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d77ff0be94849eac75e03b0d3f23800326538102e16feefc3d3e40b3ae51a098 +size 4362142880 diff --git a/model-00010-of-00062.safetensors b/model-00010-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7865aec5f71cffea660c0c2513470050ad132134 --- /dev/null +++ b/model-00010-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:98b9cc9c68ca9ee19242bd02bafda8c6c612cc6fd80d64114dfff121cc1db522 +size 4966188880 diff --git a/model-00011-of-00062.safetensors b/model-00011-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8b34e01f8e6516738893d8e67b97b4d977296dd5 --- /dev/null +++ b/model-00011-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:31c5766fb851b089719db14f1917627f125cd5d3287d1c2f138492ebc62342a0 +size 4362142872 diff --git a/model-00012-of-00062.safetensors b/model-00012-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c25f2357c931c0daba1926055eb96823d5bcc356 --- /dev/null +++ b/model-00012-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:96329b0ccd35c4b8d16d8791d062e3f58ab5239cd49c85ad92aa7fe1baf120a5 +size 4362142872 diff --git a/model-00013-of-00062.safetensors b/model-00013-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3d51242789e4625867fe51d24c99782ae05578d9 --- /dev/null +++ b/model-00013-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e0307bf0cb015cc3de8c657377e4f7d27344367053a6d3435a65179f9c776405 +size 4966188880 diff --git a/model-00014-of-00062.safetensors b/model-00014-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..30d573a8d6127baf75e4286c2a04d698129cbfd4 --- /dev/null +++ b/model-00014-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c95bc1f9cd3fffe0514d1ecc23621b0c041ac678b5aef363dae52fdffd562fd +size 4362142872 diff --git a/model-00015-of-00062.safetensors b/model-00015-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2cc658415ca2bfddb58fbe615ade5a9437bb61cd --- /dev/null +++ b/model-00015-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b8b09ec6f0a23f12f9a0ae0c6eb54162861d92ea89fdf5fa1d2a0ca0aee4477 +size 4362142872 diff --git a/model-00016-of-00062.safetensors b/model-00016-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8e20130fac8271fb10b2e12a16080e888f6a9fae --- /dev/null +++ b/model-00016-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4acc5969aad7303e894067f49053fc5f578624296dc040412a711bb04876d3b +size 4966188880 diff --git a/model-00017-of-00062.safetensors b/model-00017-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3ca2ec54f0609d34c91b568c69905745b5f6dace --- /dev/null +++ b/model-00017-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8bee1e379b2d73e6270339d30dcc815bd2295be4ddce6d539d4ba6d0659fa11c +size 4362142872 diff --git a/model-00018-of-00062.safetensors b/model-00018-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c6fd71deea8fc7403949370b69b8fa94a963ae9c --- /dev/null +++ b/model-00018-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5ad7880cae914b6841ab8e33b457eb320b2c62d70d0f0acc29a747b43f458cb2 +size 4362142872 diff --git a/model-00019-of-00062.safetensors b/model-00019-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..15e7bd7131ab977c68e24c1af5b39174e84d04d0 --- /dev/null +++ b/model-00019-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:26d6f20772c2c2e268578de8158d19799fc4c22ea5610db16a95c78fc9339aa0 +size 4966188880 diff --git a/model-00020-of-00062.safetensors b/model-00020-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9e0b2e50444047fcd7195517d2808eeda28c3437 --- /dev/null +++ b/model-00020-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c9159399cc077ab88b91b945c59a0d9c911e90d4aa386b1220260e5b4ce1d77 +size 4362142872 diff --git a/model-00021-of-00062.safetensors b/model-00021-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6509f8f9f58053747b5da4c7167582c33a726184 --- /dev/null +++ b/model-00021-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:88f19d18fe2f0979d6fe939841515af3521ef771b43aed86dddc15bedbdd2f36 +size 4362142872 diff --git a/model-00022-of-00062.safetensors b/model-00022-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8b78a65662ff1ceab70ca2d65c9c88436c1e6cb3 --- /dev/null +++ b/model-00022-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0b64dc3d0a633233b7d022c6c7f1354617fb0d4b24fb78887d33992667f69698 +size 4966188880 diff --git a/model-00023-of-00062.safetensors b/model-00023-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7afe06dc5e1f8c9605e91d3a1b7212c2fa4e5550 --- /dev/null +++ b/model-00023-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3cb20619df9745a0c07621864e01da225075959b282503f580df4d2425831c19 +size 4362142872 diff --git a/model-00024-of-00062.safetensors b/model-00024-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..685ab39b03472e00bdd153c546fb1aba84a13572 --- /dev/null +++ b/model-00024-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a953f7d8d03e5215ba16443af633dce083037a549debc712596c8eda7bd16870 +size 4362142872 diff --git a/model-00025-of-00062.safetensors b/model-00025-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..11e63ee80676cb97e47599945079f0b5c61085ee --- /dev/null +++ b/model-00025-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d4ef31c663a3dc48d159a976970ca735615783f49d0493f581810fedb1a8b50 +size 4966188880 diff --git a/model-00026-of-00062.safetensors b/model-00026-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3a8a5ab336e6257b35353279a81ed0c21b1a7560 --- /dev/null +++ b/model-00026-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c727f018923d6a01d94d1bf242d88c0eef9fbcf8e3df78f11b50275e80eb5482 +size 4362142872 diff --git a/model-00027-of-00062.safetensors b/model-00027-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bbbb926b81909e703d3c7aceddccddc8feb2d6ef --- /dev/null +++ b/model-00027-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74cf981d1ba8abfea4f48e4e65c12f34ca4dfdea64d7821453b4aa658899e9a8 +size 4362142872 diff --git a/model-00028-of-00062.safetensors b/model-00028-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9bb3cf73f53d05f347ac0df9a8cf5339caa9e648 --- /dev/null +++ b/model-00028-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a359024ee1ca5216ae1d5ca631e3d6eb67265714f9b55e178481728b45c05578 +size 4966188880 diff --git a/model-00029-of-00062.safetensors b/model-00029-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d311f33f843c74366717c2616f4e1fd10fe7775 --- /dev/null +++ b/model-00029-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:858b382360c0c57dfc17abb20d28a61e1fc0f0b34d5290383e69631c7972608c +size 4362142872 diff --git a/model-00030-of-00062.safetensors b/model-00030-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c84b32cfa66d14753b0a73ec4c61461f5d256ee4 --- /dev/null +++ b/model-00030-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8ef09d30a66af96538dc6f382129c1b297278d5baa6bb5491f926c5693274f8 +size 4362142872 diff --git a/model-00031-of-00062.safetensors b/model-00031-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cbba7195a8d0b6862da9cca44ad21a8f8f5dcf20 --- /dev/null +++ b/model-00031-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0b91e5dea235b57dfd6369137722e4a0876c6d24f4ac7177eb88c6ec1139de7 +size 4966188880 diff --git a/model-00032-of-00062.safetensors b/model-00032-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..85a1b069624eb1a4f87d50383702964f3eccc6fc --- /dev/null +++ b/model-00032-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd58845afd6b69d11744417067e73298c688a328aebf0d426ed10d997f184d3c +size 4362142872 diff --git a/model-00033-of-00062.safetensors b/model-00033-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..00d790ffa6401293df5d3072d4dea5c02f9ec2d1 --- /dev/null +++ b/model-00033-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dae2c007bcfac81441743007065e00566d7256afa3602e0c39cda361bdd97a4 +size 4362142872 diff --git a/model-00034-of-00062.safetensors b/model-00034-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3ac7358a171101c3f81f6de88b03ae4182c1e20d --- /dev/null +++ b/model-00034-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:790f5278f1e40c06876c14aec668a5fd6a24dcf8321e3b3413b992be7fffcf7e +size 4966188880 diff --git a/model-00035-of-00062.safetensors b/model-00035-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4b92f6a5d11eb3f61f8780f5247ca451b89012bf --- /dev/null +++ b/model-00035-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:02cafabad8f09dc315c9f72baba6ed4ba92cc5127c1b14e437d4e7a61bdc54b1 +size 4362142872 diff --git a/model-00036-of-00062.safetensors b/model-00036-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fb2a0b126edd091f172cdf0b271215324e9cba9c --- /dev/null +++ b/model-00036-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1742e8969ef49e8a9aa5490fbaaf868ab6b6874c9625d71074632239d04a51d1 +size 4362142872 diff --git a/model-00037-of-00062.safetensors b/model-00037-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..065a973fc147ecca4f9622cea1bf2c7bbb816b09 --- /dev/null +++ b/model-00037-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a11ce33e0ba0e2d3021d4ed63923c27f4f256c74a6a4effe0eb9102a3c1d5381 +size 4966188880 diff --git a/model-00038-of-00062.safetensors b/model-00038-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4eea8849f975934708813754a4be25bed51ceb91 --- /dev/null +++ b/model-00038-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c3153e7d6114ec71c63f99bbf33498157600b8b2ffbd3c442d2f14e37bf0f9f +size 4362142872 diff --git a/model-00039-of-00062.safetensors b/model-00039-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..364f2992e273e1ffa7964f152389257481f0ebab --- /dev/null +++ b/model-00039-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf9fce44374b00e8cd99cfc888c92768a8d210cc054de9ca8f79a49ae725b9fb +size 4362142872 diff --git a/model-00040-of-00062.safetensors b/model-00040-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..10563d88070655fadb103f74ed4d55df9faf7daa --- /dev/null +++ b/model-00040-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e65bdfb179c50adaad93e5a87855f94aad3f776f08e95ffcfe6a9d50819cf4b +size 4966188880 diff --git a/model-00041-of-00062.safetensors b/model-00041-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ece17a3f26a3d373ddf09716ec61627165476a5 --- /dev/null +++ b/model-00041-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:93cb604d3fe041f0ff181d828194498433f4a2b0518245f955f4c5666c046543 +size 4362142872 diff --git a/model-00042-of-00062.safetensors b/model-00042-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..76e58d9a1d3f38058d2e7ac4e50afe87e0b5c66a --- /dev/null +++ b/model-00042-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a50e11b046e2da47e7a8575ff1b854b26cc4ee9d364483c4f231c2c9a550d93 +size 4362142872 diff --git a/model-00043-of-00062.safetensors b/model-00043-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e6fd0785d9a494c1d139789fa799efea8a9d5b65 --- /dev/null +++ b/model-00043-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b376bf6b14b0096ff3c105f1e8bc4cf01e2186f494cfc3c87f7e2720b077cf3 +size 4966188880 diff --git a/model-00044-of-00062.safetensors b/model-00044-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9f49bb1b14b2443164fefca942882971e0748166 --- /dev/null +++ b/model-00044-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c2ca635b14164e6225e9d6d6db299f6d886149ae71a3251a72d4243fe8f2227 +size 4362142872 diff --git a/model-00045-of-00062.safetensors b/model-00045-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..058fae82d0c5473bb62832d8cf621382337bdaff --- /dev/null +++ b/model-00045-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e5300658a6c4865ad570027e4f80108173630afe8232db45d2866b06894d2b5 +size 4362142872 diff --git a/model-00046-of-00062.safetensors b/model-00046-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2011d1a9e95dbe8c8af088ce9bb566cda4b40e84 --- /dev/null +++ b/model-00046-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64106f49bd3abd1f9ec11f7290004cf80c4522a0e704481f00a2b5ac7e32f28e +size 4966188880 diff --git a/model-00047-of-00062.safetensors b/model-00047-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b78d30c966c66978dc8e5b9e15f00588d4e8e23a --- /dev/null +++ b/model-00047-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de4c06d5a4cd137ebaaf890e9c4f7feffbd0523d960fa1c0b023dc2fcf464a7c +size 4362142872 diff --git a/model-00048-of-00062.safetensors b/model-00048-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ef7d55c675147acf4bcbd3a0c26dee3f2396ba4e --- /dev/null +++ b/model-00048-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:285a83e448f134e1441d75c7b7eba25ff488be12a782b68850ba4b8e4b45b3a9 +size 4362142872 diff --git a/model-00049-of-00062.safetensors b/model-00049-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5e2ea243a17f86239d9b2bfe8d35b0af95285d53 --- /dev/null +++ b/model-00049-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5331033671e2897c7094bc785a81e6a4bbef476ff87be7c26cd659acd9edc70b +size 4966188880 diff --git a/model-00050-of-00062.safetensors b/model-00050-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e1d4669f4840d781c12b2b6cb06e413585a1ccaf --- /dev/null +++ b/model-00050-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a62381609fdc9c78aab5bbd7064c77f86cd9c5508121ca9b7f5db784ac48eac0 +size 4362142872 diff --git a/model-00051-of-00062.safetensors b/model-00051-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a3ad53c4f9fe83a10c146fee8cfd5994aea4ad38 --- /dev/null +++ b/model-00051-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9505648f7756bc6011d33aa7d10f0874f94f51cb5b764f9ada9f8f597679d3ef +size 4362142872 diff --git a/model-00052-of-00062.safetensors b/model-00052-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..02fad2b123fa8308342a829d6c260fb41689257e --- /dev/null +++ b/model-00052-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a5147309a04ec31324e651aaea26c0026526a23b2159dc295068f3cc09a05af +size 4966188880 diff --git a/model-00053-of-00062.safetensors b/model-00053-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..242c8325d8e1bcff05512161c03068c03a93a2a8 --- /dev/null +++ b/model-00053-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a6496cb0e745fed00fd4cbfc736f647a4c3b10ac49e14ed0b01956ac4372ffa +size 4362142872 diff --git a/model-00054-of-00062.safetensors b/model-00054-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a889a7a4da83c079c42d78c18b21ce5fa5547529 --- /dev/null +++ b/model-00054-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aeaa43f48931f4417425c863f7f2d9f47cc115a7698d81bc18bd630c19417382 +size 4362142872 diff --git a/model-00055-of-00062.safetensors b/model-00055-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..608643b2c81fa40c005603246437134c9b0fd4ee --- /dev/null +++ b/model-00055-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5f941e35b8d09517ac9cc7742fa83b3141e1067a63a01b8f085d270bcc849357 +size 4966188880 diff --git a/model-00056-of-00062.safetensors b/model-00056-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8a2db126be674984d60580d4e5db5fbcb7d8af79 --- /dev/null +++ b/model-00056-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ebc0fd0bf8a6c53d8ff7cb8eb529d5bf0d79db38e03eaf4395465a27d3d38bbf +size 4362142872 diff --git a/model-00057-of-00062.safetensors b/model-00057-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f2bc9ba867138fe185af77e252d9684b5caa02f2 --- /dev/null +++ b/model-00057-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c58344595817b57da0d197c22d956475c86777fd8f3c334b26a54948ea731351 +size 4362142872 diff --git a/model-00058-of-00062.safetensors b/model-00058-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..be025fadeb3eaf04d25daf09f4c04cf4fe4a5461 --- /dev/null +++ b/model-00058-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:83d63f8503d41ce82a6e36d076936b1465a492326284a8038cc6144973736d0e +size 4966188880 diff --git a/model-00059-of-00062.safetensors b/model-00059-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7b1d24e6977b2110934b7eb29071a4a1b6e2ac85 --- /dev/null +++ b/model-00059-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6c8f6d78dcfa3cfb6d7859e2efd194238062b279a705c22e58198e0bbf0cb821 +size 4362142872 diff --git a/model-00060-of-00062.safetensors b/model-00060-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b9b3c04f5f44ef52714bbbcfdffb19d708eb40bc --- /dev/null +++ b/model-00060-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b4004856358e8b24cad3485dacec86d14ae1fff08921c00222960a592cf2e580 +size 4362142872 diff --git a/model-00061-of-00062.safetensors b/model-00061-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ca115699148c92f445a8b99d97f197263a414202 --- /dev/null +++ b/model-00061-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5c5e6965aa2233410d5fcee3825f7e4c111bb2a2b489d99dd6f54527dfb8487 +size 4362241496 diff --git a/model-00062-of-00062.safetensors b/model-00062-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2762bb1914af0ec0b351a5f478c666c7efbf09db --- /dev/null +++ b/model-00062-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a9e9d49177e91909b535d578b7fc969bb0c295342548b2755c683643be58efbc +size 4202692736 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..8aab69e95f0d7da106bad45a4e7c35897ecf3d80 --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,730 @@ +{ + "metadata": { + "total_size": 282214825984 + }, + "weight_map": { + "lm_head.weight": "model-00062-of-00062.safetensors", + "model.embed_tokens.weight": "model-00001-of-00062.safetensors", + "model.layers.0.input_layernorm.weight": "model-00002-of-00062.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00002-of-00062.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.1.input_layernorm.weight": "model-00003-of-00062.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00003-of-00062.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.10.input_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.11.input_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.input_layernorm.weight": "model-00011-of-00062.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00011-of-00062.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.13.input_layernorm.weight": "model-00012-of-00062.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00012-of-00062.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.14.input_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.15.input_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.input_layernorm.weight": "model-00014-of-00062.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00014-of-00062.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.17.input_layernorm.weight": "model-00015-of-00062.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00015-of-00062.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.18.input_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.19.input_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.2.input_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.20.input_layernorm.weight": "model-00017-of-00062.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00017-of-00062.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.21.input_layernorm.weight": "model-00018-of-00062.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00018-of-00062.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.22.input_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.23.input_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.input_layernorm.weight": "model-00020-of-00062.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00020-of-00062.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.25.input_layernorm.weight": "model-00021-of-00062.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00021-of-00062.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.26.input_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.27.input_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.input_layernorm.weight": "model-00023-of-00062.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.28.mlp.gate_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.28.mlp.up_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00023-of-00062.safetensors", + "model.layers.28.self_attn.k_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.self_attn.q_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.self_attn.v_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.29.input_layernorm.weight": "model-00024-of-00062.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.29.mlp.gate_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.mlp.up_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00024-of-00062.safetensors", + "model.layers.29.self_attn.k_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.self_attn.q_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.self_attn.v_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.3.input_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.30.input_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.30.mlp.gate_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.mlp.up_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.30.self_attn.k_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.self_attn.q_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.self_attn.v_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.31.input_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.mlp.gate_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.mlp.up_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.k_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.q_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.v_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.input_layernorm.weight": "model-00026-of-00062.safetensors", + "model.layers.32.mlp.down_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.32.mlp.gate_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.32.mlp.up_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.32.post_attention_layernorm.weight": "model-00026-of-00062.safetensors", + "model.layers.32.self_attn.k_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.self_attn.o_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.self_attn.q_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.self_attn.v_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.33.input_layernorm.weight": "model-00027-of-00062.safetensors", + "model.layers.33.mlp.down_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.33.mlp.gate_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.mlp.up_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.33.post_attention_layernorm.weight": "model-00027-of-00062.safetensors", + "model.layers.33.self_attn.k_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.self_attn.o_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.self_attn.q_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.self_attn.v_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.34.input_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.34.mlp.down_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.34.mlp.gate_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.mlp.up_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.post_attention_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.34.self_attn.k_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.self_attn.o_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.self_attn.q_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.self_attn.v_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.35.input_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.35.mlp.down_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.mlp.gate_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.mlp.up_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.post_attention_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.k_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.o_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.q_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.v_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.input_layernorm.weight": "model-00029-of-00062.safetensors", + "model.layers.36.mlp.down_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.36.mlp.gate_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.36.mlp.up_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.36.post_attention_layernorm.weight": "model-00029-of-00062.safetensors", + "model.layers.36.self_attn.k_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.self_attn.o_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.self_attn.q_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.self_attn.v_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.37.input_layernorm.weight": "model-00030-of-00062.safetensors", + "model.layers.37.mlp.down_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.37.mlp.gate_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.mlp.up_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.37.post_attention_layernorm.weight": "model-00030-of-00062.safetensors", + "model.layers.37.self_attn.k_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.self_attn.o_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.self_attn.q_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.self_attn.v_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.38.input_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.38.mlp.down_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.38.mlp.gate_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.mlp.up_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.post_attention_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.38.self_attn.k_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.self_attn.o_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.self_attn.q_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.self_attn.v_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.39.input_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.39.mlp.down_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.mlp.gate_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.mlp.up_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.post_attention_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.k_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.o_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.q_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.v_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.4.input_layernorm.weight": "model-00005-of-00062.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00005-of-00062.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.40.input_layernorm.weight": "model-00032-of-00062.safetensors", + "model.layers.40.mlp.down_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.40.mlp.gate_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.40.mlp.up_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.40.post_attention_layernorm.weight": "model-00032-of-00062.safetensors", + "model.layers.40.self_attn.k_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.40.self_attn.o_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.40.self_attn.q_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.40.self_attn.v_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.41.input_layernorm.weight": "model-00033-of-00062.safetensors", + "model.layers.41.mlp.down_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.41.mlp.gate_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.mlp.up_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.41.post_attention_layernorm.weight": "model-00033-of-00062.safetensors", + "model.layers.41.self_attn.k_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.self_attn.o_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.self_attn.q_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.self_attn.v_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.42.input_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.42.mlp.down_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.42.mlp.gate_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.mlp.up_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.post_attention_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.42.self_attn.k_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.self_attn.o_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.self_attn.q_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.self_attn.v_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.43.input_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.43.mlp.down_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.mlp.gate_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.mlp.up_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.post_attention_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.k_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.o_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.q_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.v_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.input_layernorm.weight": "model-00035-of-00062.safetensors", + "model.layers.44.mlp.down_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.44.mlp.gate_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.44.mlp.up_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.44.post_attention_layernorm.weight": "model-00035-of-00062.safetensors", + "model.layers.44.self_attn.k_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.self_attn.o_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.self_attn.q_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.self_attn.v_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.45.input_layernorm.weight": "model-00036-of-00062.safetensors", + "model.layers.45.mlp.down_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.45.mlp.gate_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.mlp.up_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.45.post_attention_layernorm.weight": "model-00036-of-00062.safetensors", + "model.layers.45.self_attn.k_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.self_attn.o_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.self_attn.q_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.self_attn.v_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.46.input_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.46.mlp.down_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.46.mlp.gate_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.mlp.up_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.post_attention_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.46.self_attn.k_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.self_attn.o_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.self_attn.q_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.self_attn.v_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.47.input_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.47.mlp.down_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.mlp.gate_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.mlp.up_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.post_attention_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.k_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.o_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.q_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.v_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.input_layernorm.weight": "model-00038-of-00062.safetensors", + "model.layers.48.mlp.down_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.48.mlp.gate_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.48.mlp.up_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.48.post_attention_layernorm.weight": "model-00038-of-00062.safetensors", + "model.layers.48.self_attn.k_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.self_attn.o_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.self_attn.q_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.self_attn.v_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.49.input_layernorm.weight": "model-00039-of-00062.safetensors", + "model.layers.49.mlp.down_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.49.mlp.gate_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.mlp.up_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.49.post_attention_layernorm.weight": "model-00039-of-00062.safetensors", + "model.layers.49.self_attn.k_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.self_attn.o_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.self_attn.q_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.self_attn.v_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.5.input_layernorm.weight": "model-00006-of-00062.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00006-of-00062.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.50.input_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.50.mlp.down_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.50.mlp.gate_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.mlp.up_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.post_attention_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.50.self_attn.k_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.self_attn.o_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.self_attn.q_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.self_attn.v_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.51.input_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.51.mlp.down_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.mlp.gate_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.mlp.up_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.post_attention_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.k_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.o_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.q_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.v_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.input_layernorm.weight": "model-00041-of-00062.safetensors", + "model.layers.52.mlp.down_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.52.mlp.gate_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.52.mlp.up_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.52.post_attention_layernorm.weight": "model-00041-of-00062.safetensors", + "model.layers.52.self_attn.k_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.self_attn.o_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.self_attn.q_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.self_attn.v_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.53.input_layernorm.weight": "model-00042-of-00062.safetensors", + "model.layers.53.mlp.down_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.53.mlp.gate_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.mlp.up_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.53.post_attention_layernorm.weight": "model-00042-of-00062.safetensors", + "model.layers.53.self_attn.k_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.self_attn.o_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.self_attn.q_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.self_attn.v_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.54.input_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.54.mlp.down_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.54.mlp.gate_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.mlp.up_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.post_attention_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.54.self_attn.k_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.self_attn.o_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.self_attn.q_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.self_attn.v_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.55.input_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.55.mlp.down_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.mlp.gate_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.mlp.up_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.post_attention_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.k_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.o_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.q_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.v_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.input_layernorm.weight": "model-00044-of-00062.safetensors", + "model.layers.56.mlp.down_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.56.mlp.gate_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.56.mlp.up_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.56.post_attention_layernorm.weight": "model-00044-of-00062.safetensors", + "model.layers.56.self_attn.k_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.self_attn.o_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.self_attn.q_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.self_attn.v_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.57.input_layernorm.weight": "model-00045-of-00062.safetensors", + "model.layers.57.mlp.down_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.57.mlp.gate_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.mlp.up_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.57.post_attention_layernorm.weight": "model-00045-of-00062.safetensors", + "model.layers.57.self_attn.k_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.self_attn.o_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.self_attn.q_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.self_attn.v_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.58.input_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.58.mlp.down_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.58.mlp.gate_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.mlp.up_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.post_attention_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.58.self_attn.k_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.self_attn.o_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.self_attn.q_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.self_attn.v_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.59.input_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.59.mlp.down_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.mlp.gate_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.mlp.up_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.post_attention_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.k_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.o_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.q_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.v_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.6.input_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.60.input_layernorm.weight": "model-00047-of-00062.safetensors", + "model.layers.60.mlp.down_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.60.mlp.gate_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.60.mlp.up_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.60.post_attention_layernorm.weight": "model-00047-of-00062.safetensors", + "model.layers.60.self_attn.k_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.60.self_attn.o_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.60.self_attn.q_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.60.self_attn.v_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.61.input_layernorm.weight": "model-00048-of-00062.safetensors", + "model.layers.61.mlp.down_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.61.mlp.gate_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.mlp.up_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.61.post_attention_layernorm.weight": "model-00048-of-00062.safetensors", + "model.layers.61.self_attn.k_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.self_attn.o_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.self_attn.q_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.self_attn.v_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.62.input_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.62.mlp.down_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.62.mlp.gate_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.mlp.up_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.post_attention_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.62.self_attn.k_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.self_attn.o_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.self_attn.q_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.self_attn.v_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.63.input_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.63.mlp.down_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.mlp.gate_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.mlp.up_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.post_attention_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.k_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.o_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.q_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.v_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.input_layernorm.weight": "model-00050-of-00062.safetensors", + "model.layers.64.mlp.down_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.64.mlp.gate_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.64.mlp.up_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.64.post_attention_layernorm.weight": "model-00050-of-00062.safetensors", + "model.layers.64.self_attn.k_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.self_attn.o_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.self_attn.q_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.self_attn.v_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.65.input_layernorm.weight": "model-00051-of-00062.safetensors", + "model.layers.65.mlp.down_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.65.mlp.gate_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.mlp.up_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.65.post_attention_layernorm.weight": "model-00051-of-00062.safetensors", + "model.layers.65.self_attn.k_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.self_attn.o_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.self_attn.q_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.self_attn.v_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.66.input_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.66.mlp.down_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.66.mlp.gate_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.mlp.up_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.post_attention_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.66.self_attn.k_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.self_attn.o_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.self_attn.q_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.self_attn.v_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.67.input_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.67.mlp.down_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.mlp.gate_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.mlp.up_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.post_attention_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.k_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.o_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.q_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.v_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.input_layernorm.weight": "model-00053-of-00062.safetensors", + "model.layers.68.mlp.down_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.68.mlp.gate_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.68.mlp.up_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.68.post_attention_layernorm.weight": "model-00053-of-00062.safetensors", + "model.layers.68.self_attn.k_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.self_attn.o_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.self_attn.q_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.self_attn.v_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.69.input_layernorm.weight": "model-00054-of-00062.safetensors", + "model.layers.69.mlp.down_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.69.mlp.gate_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.mlp.up_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.69.post_attention_layernorm.weight": "model-00054-of-00062.safetensors", + "model.layers.69.self_attn.k_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.self_attn.o_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.self_attn.q_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.self_attn.v_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.7.input_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.70.input_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.70.mlp.down_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.70.mlp.gate_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.mlp.up_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.post_attention_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.70.self_attn.k_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.self_attn.o_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.self_attn.q_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.self_attn.v_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.71.input_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.71.mlp.down_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.mlp.gate_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.mlp.up_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.post_attention_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.k_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.o_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.q_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.v_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.input_layernorm.weight": "model-00056-of-00062.safetensors", + "model.layers.72.mlp.down_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.72.mlp.gate_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.72.mlp.up_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.72.post_attention_layernorm.weight": "model-00056-of-00062.safetensors", + "model.layers.72.self_attn.k_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.self_attn.o_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.self_attn.q_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.self_attn.v_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.73.input_layernorm.weight": "model-00057-of-00062.safetensors", + "model.layers.73.mlp.down_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.73.mlp.gate_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.mlp.up_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.73.post_attention_layernorm.weight": "model-00057-of-00062.safetensors", + "model.layers.73.self_attn.k_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.self_attn.o_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.self_attn.q_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.self_attn.v_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.74.input_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.74.mlp.down_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.74.mlp.gate_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.mlp.up_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.post_attention_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.74.self_attn.k_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.self_attn.o_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.self_attn.q_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.self_attn.v_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.75.input_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.75.mlp.down_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.mlp.gate_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.mlp.up_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.post_attention_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.k_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.o_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.q_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.v_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.input_layernorm.weight": "model-00059-of-00062.safetensors", + "model.layers.76.mlp.down_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.76.mlp.gate_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.76.mlp.up_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.76.post_attention_layernorm.weight": "model-00059-of-00062.safetensors", + "model.layers.76.self_attn.k_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.self_attn.o_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.self_attn.q_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.self_attn.v_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.77.input_layernorm.weight": "model-00060-of-00062.safetensors", + "model.layers.77.mlp.down_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.77.mlp.gate_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.mlp.up_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.77.post_attention_layernorm.weight": "model-00060-of-00062.safetensors", + "model.layers.77.self_attn.k_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.self_attn.o_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.self_attn.q_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.self_attn.v_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.78.input_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.78.mlp.down_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.78.mlp.gate_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.mlp.up_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.post_attention_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.78.self_attn.k_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.self_attn.o_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.self_attn.q_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.self_attn.v_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.79.input_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.79.mlp.down_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.mlp.gate_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.mlp.up_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.post_attention_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.k_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.o_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.q_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.v_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.8.input_layernorm.weight": "model-00008-of-00062.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00008-of-00062.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.9.input_layernorm.weight": "model-00009-of-00062.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00009-of-00062.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00008-of-00062.safetensors", + "model.norm.weight": "model-00061-of-00062.safetensors" + } +}