Replace model with mpt-7b-instruct, loaded in f16 and sharded to 2GB chunks

config.json

@@ -1,16 +1,16 @@
 {
-  "_name_or_path": "mosaicml/mpt-7b-storywriter",
+  "_name_or_path": "mosaicml/mpt-7b-instruct",
   "architectures": [
     "MPTForCausalLM"
   ],
   "attn_config": {
     "alibi": true,
-    "alibi_bias_max": 16,
+    "alibi_bias_max": 8,
     "attn_impl": "torch",
     "attn_pdrop": 0,
     "attn_type": "multihead_attention",
     "attn_uses_sequence_id": false,
-    "clip_qkv": 6,
+    "clip_qkv": null,
     "prefix_lm": false,
     "qk_ln": false,
     "softmax_scale": null
@@ -37,7 +37,7 @@
   "init_device": "cpu",
   "learned_pos_emb": true,
   "logit_scale": null,
-  "max_seq_len": 65536,
+  "max_seq_len": 2048,
   "model_type": "mpt",
   "n_heads": 32,
   "n_layers": 32,
@@ -46,7 +46,7 @@
   "resid_pdrop": 0,
   "tokenizer_name": "EleutherAI/gpt-neox-20b",
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.28.1",
+  "transformers_version": "4.29.2",
   "use_cache": true,
   "verbose": 0,
   "vocab_size": 50432

generation_config.json

@@ -1,13 +1,5 @@
 {
-  "bos_token_id":0,
-  "_from_model_config":false,
-  "do_sample":true,
-  "eos_token_id":0,
-  "max_new_tokens":512,
-  "no_repeat_ngram_size":10,
-  "pad_token_id":0,
-  "penalty_alpha":0.5,
-  "repetition_penalty":1.05,
-  "top_k":4,
-  "transformers_version":"4.28.1"
-}
+  "_from_model_config": true,
+  "transformers_version": "4.29.2",
+  "use_cache": false
+}

pytorch_model-00001-of-00007.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e06e21e5597362475c4647f78111969186a50b13e842c704f6f2aa4a4891ad16
-size 1889607551
+oid sha256:504b9923f3ed1c8d9605c2fe470db548a5952ca8c488ee8c3b87f5ba3c8e3e88
+size 1889608022

pytorch_model-00002-of-00007.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e9e695959203c11763b1d19c9441279afccee5b0259099f6153d12eff1e4e11b
-size 1879139801
+oid sha256:dc128efcdc9cfd2d00e88ad1686996658e0a7c53a35281bcbcfd248c09584bf4
+size 1879140272

pytorch_model-00003-of-00007.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4b686f5548dd7b6e385bc4d6960edc519fc7be06070e62fd5cc8582841d109c7
-size 1979794567
+oid sha256:f13268b4cd1ae32acc89155a3f09da882951ea7b254aece34693b47986b97a68
+size 1979795038

pytorch_model-00004-of-00007.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a24a2d01c2654329961ef0479653b571be6cd3702e5398cee93af4ed7dad995
-size 1912694233
+oid sha256:36bcd56375703ea98f8e7fd571bb2188bbf1b140d39660e49bafa10b233edc6f
+size 1912694704

pytorch_model-00005-of-00007.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:90425be6ec1c91e57316858f3bc4b7496863e4d938cc181b0fe7722d8f88051b
-size 1879131271
+oid sha256:9d30513408b88b1a77bf1c5b88bf285751e9e163bf3866239da771c18a74d642
+size 1879131806

pytorch_model-00006-of-00007.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a32c8295b8c663a339e6375cae8bf585376bdc7898d6ec3eac3aa9533ccf4bbe
-size 1879139801
+oid sha256:37653de6fb3ed72181e55e8835a610f22b74254ff9c8c3bbc2d3a801615fc3e1
+size 1879140272

pytorch_model-00007-of-00007.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5fb7cbcc874e2dcba3e928017de4d60db2a2e4a1a32090a49aec68ab0a252d0b
-size 1879130997
+oid sha256:435e5dff52b384064a2984eafe651c9aaabe466a16dda239b60cccbb4770e331
+size 1879131468