Autobots's picture

23 1

Autobots

autobots

·

AI & ML interests

Chatbots

Organizations

None yet

autobots's activity

New activity in maddes8cht/tiiuae-falcon-40b-instruct-gguf over 1 year ago

Failure loading in text-generation-webui.

#1 opened over 1 year ago by

New activity in autobots/Nyanners-Narrator-TTS-Tortoise over 1 year ago

How to load the voice model?

#1 opened over 1 year ago by

New activity in TheBloke/fiction.live-Kimiko-V2-70B-GGUF over 1 year ago

Repetition issues?

#1 opened over 1 year ago by

New activity in CalderaAI/30B-Epsilon over 1 year ago

Is this the 'better than Lazarus' model?

#1 opened over 1 year ago by

7erminalVelociraptor

New activity in reeducator/vicuna-13b-cocktail over 1 year ago

Cocktail Testing and Discussion

#1 opened almost 2 years ago by deleted

New activity in notstoic/OPT-13B-Erebus-4bit-128g almost 2 years ago

I get expected Float found Half error

#1 opened almost 2 years ago by

New activity in TheBloke/OpenAssistant-SFT-7-Llama-30B-GPTQ almost 2 years ago

RuntimeError: shape '[-1, 1024, 6656]' is invalid for input of size 44302336

#6 opened almost 2 years ago by

New activity in reeducator/bluemoonrp-13b almost 2 years ago

Initial testing (epoch 3)

#1 opened almost 2 years ago by deleted

New activity in reeducator/vicuna-13b-free almost 2 years ago

text-generation-webui: AttributeError: 'Offload_LlamaModel' object has no attribute 'preload', when trying to generate text

#21 opened almost 2 years ago by

New activity in TheBloke/alpaca-lora-65B-GPTQ almost 2 years ago

Beware! Requires new cuda.

#5 opened almost 2 years ago by

New activity in maderix/llama-65b-4bit almost 2 years ago

This is a v1 model. Doesn't run on newer GPTQ anymore.

#6 opened almost 2 years ago by

New activity in reeducator/vicuna-13b-free almost 2 years ago

You may want to add an "act-order" GPTQ quantization.

#22 opened almost 2 years ago by

New activity in autobots/gpt-j-fourchannel-4bit almost 2 years ago

ggml-gpt-fourchannel-q4_0.bin does not work on llama.cpp

#1 opened almost 2 years ago by

FenixInDarkSolo

New activity in TheBloke/wizardLM-7B-HF almost 2 years ago

This config also has cache disabled.

#1 opened almost 2 years ago by

New activity in autobots/alpaca-7b-native-enhanced-4bit almost 2 years ago

conig.json

#1 opened almost 2 years ago by

New activity in notstoic/OPT-13B-Erebus-4bit-128g almost 2 years ago

Set "use_cache": true for a speedup.

#4 opened almost 2 years ago by

New activity in teknium/GPT4-x-Alpaca13b-RolePlayLora-4bit-v2 almost 2 years ago

This is made with the larger lora?

#2 opened almost 2 years ago by

New activity in 8bit-coder/alpaca-7b-nativeEnhanced almost 2 years ago

Loading the model 26gb?

#7 opened almost 2 years ago by

Can we get the GPTQ quantized model?

#5 opened almost 2 years ago by

New activity in notstoic/OPT-13B-Erebus-4bit-128g almost 2 years ago

Can you encode this without group size?

#3 opened almost 2 years ago by