This is a merge of pre-trained language models created using mergekit.
UpStage's conclusionary limitations of their research:
"Our study on the Depth Up-Scaling (DUS) has important limitations and considerations. One key limitation is the need for more thorough explorations of hyperparameters used in the DUS approach. Namely, we removed m = 8 layers from both ends of our base model, primarily due to hardware limitations. However, we have not yet determined if this value is optimal for enhancing performance. The extended time and cost of continued pretraining made it challenging to conduct more comprehensive experiments, which we aim to address in future work through various comparative analyses."
This model was made to help test whether 10.7B parameters (m = 8) is better or worse than 10.7B+ parameters (m < 8)
Merge Details
Merge Method
This model was merged using the passthrough merge method.
The following YAML configuration was used to produce this model:
yaml
1dtype: bfloat16
2merge_method: passthrough
3# Depth UpScaled (DUS) version of Mistral-7B-Instruct-v0.24# where m = 4 (The number of layers to remove from the model)5# s = 56 (The number of layers the model will have after the DUS)6slices:7-sources:8-layer_range:[0,28]9model: /Users/jsarnecki/opt/Workspace/mistralai/Mistral-7B-Instruct-v0.2
10-sources:11-layer_range:[4,32]12model: /Users/jsarnecki/opt/Workspace/mistralai/Mistral-7B-Instruct-v0.2
13
Model Card for Mistral-7B-Instruct-v0.2
The Mistral-7B-Instruct-v0.2 Large Language Model (LLM) is an instruct fine-tuned version of the Mistral-7B-v0.2.
Mistral-7B-v0.2 has the following changes compared to Mistral-7B-v0.1
In order to leverage instruction fine-tuning, your prompt should be surrounded by [INST] and [/INST] tokens. The very first instruction should begin with a begin of sentence id. The next instructions should not. The assistant generation will be ended by the end-of-sentence token id.
E.g.
text = "<s>[INST] What is your favourite condiment? [/INST]"
"Well, I'm quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I'm cooking up in the kitchen!</s> "
"[INST] Do you have mayonnaise recipes? [/INST]"
This format is available as a chat template via the apply_chat_template() method:
python
1from transformers import AutoModelForCausalLM, AutoTokenizer
23device ="cuda"# the device to load the model onto45model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")6tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")78messages =[9{"role":"user","content":"What is your favourite condiment?"},10{"role":"assistant","content":"Well, I'm quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I'm cooking up in the kitchen!"},11{"role":"user","content":"Do you have mayonnaise recipes?"}12]1314encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")1516model_inputs = encodeds.to(device)17model.to(device)1819generated_ids = model.generate(model_inputs, max_new_tokens=1000, do_sample=True)20decoded = tokenizer.batch_decode(generated_ids)21print(decoded[0])
Troubleshooting
If you see the following error:
Traceback (most recent call last):
File "", line 1, in
File "/transformers/models/auto/auto_factory.py", line 482, in from_pretrained
config, kwargs = AutoConfig.from_pretrained(
File "/transformers/models/auto/configuration_auto.py", line 1022, in from_pretrained
config_class = CONFIG_MAPPING[config_dict["model_type"]]
File "/transformers/models/auto/configuration_auto.py", line 723, in getitem
raise KeyError(key)
KeyError: 'mistral'
This should not be required after transformers-v4.33.4.
Limitations
The Mistral 7B Instruct model is a quick demonstration that the base model can be easily fine-tuned to achieve compelling performance.
It does not have any moderation mechanisms. We're looking forward to engaging with the community on ways to
make the model finely respect guardrails, allowing for deployment in environments requiring moderated outputs.
The Mistral AI Team
Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Blanche Savary, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Emma Bou Hanna, Florian Bressand, Gianna Lengyel, Guillaume Bour, Guillaume Lample, Lélio Renard Lavaud, Louis Ternon, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Théophile Gervet, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed.
