add 0.2 revision with new weights

README.md

@@ -38,6 +38,11 @@ Applied open Sonnet datasets containing ~1.2mn question/answer pairs for fine-tu
 <|begin_of_text|>{prompt}
 ```
 
+## Release History
+
+ * v0.1 - [2025-02-08] Initial release, trained to 512 steps
+ * v0.2 - [2025-02-10] Restarted training with cleaner dataset, ran to 1024 steps
+
 ## Credits
 
-Thanks to Meta and [mlfoundations-dev](https://huggingface.co/mlfoundations-dev) for providing the data used to create this fine-tuning.
+Thanks to Meta and [mlfoundations-dev](https://huggingface.co/mlfoundations-dev) for providing the data used to create this fine-tuning.

config.json

@@ -30,9 +30,9 @@
   },
   "rope_theta": 500000.0,
   "tie_word_embeddings": false,
-  "torch_dtype": "bfloat16",
-  "transformers_version": "4.48.2",
-  "unsloth_version": "2025.2.5",
+  "torch_dtype": "float16",
+  "transformers_version": "4.46.0",
+  "unsloth_version": "2025.2.4",
   "use_cache": true,
   "vocab_size": 128256
 }

generation_config.json

@@ -7,5 +7,5 @@
   "pad_token_id": 128004,
   "temperature": 0.6,
   "top_p": 0.9,
-  "transformers_version": "4.48.2"
+  "transformers_version": "4.46.0"
 }

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:06039f89149eb2298e0499e727e5eb3afe98ee6b7bc5914c2622962ab5bb5698
-size 4976698672
+oid sha256:15960273496240854588c39b46d1b703b080d4bbd067c8801f2e12bc121347b7
+size 4976698592

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e6dca2e09743e4b1b621c190ba82ae22e1e10d708bc8182f2d952121124b88ae
-size 4999802720
+oid sha256:685323acd4d9159f51d7807d5c14001aba04232a5d581e7350effdcaa9a2dd8d
+size 4999802616

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:133e869b4d17caea5da38e929470f0809352caa8a4b07f66084156d52a5f6418
-size 4915916176
+oid sha256:381a574d81c9a926da69f4ee63e2945454161600048128b73c3b0e0b1f3681e7
+size 4915916080

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:073ae6cfd40d18ac769bc6b27b60b819e72281dd3b73ab29197fa7fbda931b05
+oid sha256:7d4d7b88162c2ff2306f622b8f826737907e1f1d3cc58eacf854bc30e35cf905
 size 1168138808

tokenizer_config.json

@@ -2050,10 +2050,9 @@
     }
   },
   "bos_token": "<|begin_of_text|>",
-  "chat_template": "{% if 'role' in messages[0] %}{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ messages[0]['content'] + '\n\n' }}{% set loop_messages = messages[1:] %}{% else %}{{ 'Below are some instructions that describe some tasks. Write responses that appropriately complete each request.' + '\n\n' }}{% set loop_messages = messages %}{% endif %}{% for message in loop_messages %}{% if message['role'] == 'user' %}{{ '### Instruction:\n' + message['content'] + '\n\n' }}{% elif message['role'] == 'assistant' %}{{ '### Response:\n' + message['content'] + eos_token + '\n\n' }}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '### Response:\n' }}{% endif %}{% else %}{{ bos_token }}{% if messages[0]['from'] == 'system' %}{{ messages[0]['value'] + '\n\n' }}{% set loop_messages = messages[1:] %}{% else %}{{ 'Below are some instructions that describe some tasks. Write responses that appropriately complete each request.' + '\n\n' }}{% set loop_messages = messages %}{% endif %}{% for message in loop_messages %}{% if message['from'] == 'human' %}{{ '### Instruction:\n' + message['value'] + '\n\n' }}{% elif message['from'] == 'gpt' %}{{ '### Response:\n' + message['value'] + eos_token + '\n\n' }}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '### Response:\n' }}{% endif %}{% endif %}",
+  "chat_template": "{% if 'role' in messages[0] %}{% if messages[0]['role'] == 'system' %}{% if messages[1]['role'] == 'user' %}{{ bos_token + '[INST] <<SYS>>\n' + messages[0]['content'] + '\n<</SYS>>\n\n' + messages[1]['content'] + ' [/INST]' }}{% set loop_messages = messages[2:] %}{% else %}{{ bos_token + '[INST] ' + messages[0]['content'] + ' [/INST]' }}{% set loop_messages = messages[1:] %}{% endif %}{% else %}{% set loop_messages = messages %}{% endif %}{% for message in loop_messages %}{% if message['role'] == 'user' %}{{ bos_token + '[INST] ' + message['content'].strip() + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ ' ' + message['content'].strip() + ' ' + eos_token }}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}{% else %}{% if messages[0]['from'] == 'system' %}{% if messages[1]['from'] == 'human' %}{{ bos_token + '[INST] <<SYS>>\n' + messages[0]['value'] + '\n<</SYS>>\n\n' + messages[1]['value'] + ' [/INST]' }}{% set loop_messages = messages[2:] %}{% else %}{{ bos_token + '[INST] ' + messages[0]['value'] + ' [/INST]' }}{% set loop_messages = messages[1:] %}{% endif %}{% else %}{% set loop_messages = messages %}{% endif %}{% for message in loop_messages %}{% if message['from'] == 'human' %}{{ bos_token + '[INST] ' + message['value'].strip() + ' [/INST]' }}{% elif message['from'] == 'gpt' %}{{ ' ' + message['value'].strip() + ' ' + eos_token }}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}{% endif %}",
   "clean_up_tokenization_spaces": true,
   "eos_token": "<|end_of_text|>",
-  "extra_special_tokens": {},
   "model_input_names": [
     "input_ids",
     "attention_mask"