… but the 32,000 tokens per minute presumably prevents free API calls from using more than 32,000 tokens? If so max cost would be 22 cents per call for input, so a lot less expensive overall
CODE
-
Python NumPy for Data Science and Big Data Applications
By
–
Python for #DataScience: NumPy
by Sawsan Yusuf @Medium Read more: https://
buff.ly/3THcjUj #BigData #ArtificialIntelligence #MI #AI cc: @pbalakrishnarao @ravikikan @kuriharan -
Mixtral-8x22B Fine-tuning Configuration with LoRA Parameters
By
–
base_model: mistral-community/Mixtral-8x22B-v0.1
model_type: AutoModelForCausalLM
tokenizer_type: LlamaTokenizer
trust_remote_code: true load_in_8bit: false
load_in_4bit: false
strict: false unfrozen_parameters: – ^lm_head.weight$ – ^model.embed_tokens.weight$ – -
Selective Layer Freezing for Language Model Fine-tuning Strategy
By
–
Freezing all layers except these: – ^lm_head.weight$ – ^model.embed_tokens.weight$ – model.layers.4[4-9]+.block_sparse_moe.gate – model.layers.4[4-9]+.block_sparse_moe.experts – model.layers.5[0-5]+.block_sparse_moe.gate – model.layers.5[0-5]+.block_sparse_moe.experts
-
Fair Testing: Providing Devin with Complete Files and Guidance
By
–
realized the last 2 days have been completely unfair to devin – i didnt reupload the .py thing it is supposed to clone so it was just churning around. now i gave it both files + gave a methodological hint.
-
CUDA kernels for GPT-2 forward pass implementation in llm.c
By
–
Okay I did a first quick pass of naive CUDA kernels for the forward pass of GPT-2 and pushed everything to one file in llm.c, Still only ~1000 lines of code: https://
github.com/karpathy/llm.c
/blob/master/train_gpt2.cu
… Current per iteration timings on my Lambda box <3 A100 40GB PCIe, B=4, T=1024:
– llm.c: 111ms
– -
GPT-5 Agents Coming Sooner (12 words max)
By
–
GPT-5 Agents are Coming Sooner Than You Think – OpenAI is already testing early versions of GPT-5 with enterprise clients – GPT-5 will have the ability to call AI agents to perform tasks autonomously – Agents give businesses an unfair advantage, as demonstrated by a study at
-
Popular Use Cases for AI Agents
By
–
Popular Use Cases for AI Agents – Automatic research agents running 24/7 – Software engineering, building, optimizing, and debugging code – Customer service, with AI agents handling a significant portion of conversations
-
GPT-4 Turbo Latest Version Improves Code Generation Efficiency
By
–
Side-by-side comparison between the latest version of gpt-4-turbo and the previous one, 0125-preview.
— Pietro Schirano (@skirano) 10 avril 2024
Not only is the new version less verbose and it goes directly into code, but it also (rightfully so) decides to add a flag to download the highest quality video.
Smart! 👨🏫 pic.twitter.com/eQdQ6Xg1sPSide-by-side comparison between the latest version of gpt-4-turbo and the previous one, 0125-preview. Not only is the new version less verbose and it goes directly into code, but it also (rightfully so) decides to add a flag to download the highest quality video. Smart!
-
Weight-Only Quantization and Security Wrapper Integration for LLMs
By
–
PenghuiCheng for integrating weight-only quantization to accelerate transformer-based models on @Intel platforms
https://
python.langchain.com/docs/integrati
ons/llms/weight_only_quantization/
… And JamsheedMistri for adding the [
@layerup_
](
https://
x.com/layerup_) security wrapper for LLMs
https://
python.langchain.com/docs/integrati
ons/llms/layerup_security/
… (5/n)