cmp-nct / ggllm.cpp

Falcon LLM ggml framework with CPU and GPU support

cmp-nct/ggllm.cpp Issues

Metal support
Updated 7 months ago2
Falcon quantization lets you choose q4_k_m, etc but you just get q4_k
Closed 8 months ago7
#1 performance requirement
Updated 10 months ago
Upcoming PR - Pushing the Context limit to 8k+ for all existing Falcon models - Longrange Falcon flights
Updated 10 months ago5
Performance - heads up
Updated 10 months ago2
--help , pipes and inconsistent help text
Updated a year ago7
linking error with static build
Updated a year ago2
Unable to run TheBloke Falcon40b-instruct
Updated a year ago31
Add support for AMX instructions (bf16 and/or int8)
Updated a year ago
Log the version of cuda that is being used
Updated a year ago
Is there any GUI or Web UI for ggllm.cpp?
Updated a year ago13
Would ggllm.cpp adopt the upcoming ggml fileformat (GGUF)?
Updated a year ago1
Even in interactive mode, multiturn conversation is not possible.
Updated a year ago3
Can't falcon_convert on OpenBuddy Falcon 7B model, KeyError [fixed]
Updated a year ago1
Debug Timings No Longer Working
Closed a year ago2
Instruct Mode Issue
Closed a year ago5
Could we also get the Makefile updated to build a libfalcon.so
Updated a year ago3
K Quant 64 support - quite a feat to integrate
Updated a year ago4
Performance at high context (18k+)
Updated a year ago4
OpenBLAS and CLBlast support
Updated a year ago4
Does the prompt cache work? I got an alignment assert when I turned it on.
Updated a year ago19
Can you migrate the new server from llama.cpp to here?
Updated a year ago3
release builds should have other names than "llama-master-codestring.zip"
Closed a year ago2
Steps forward - Tokenizer
Updated a year ago14
Something weird is going on with -ngl
Closed a year ago12
Apple Silicon Unable To Build
Closed a year ago6
Mul_mat Speedup??
Updated a year ago12
Implement ChatGLM.cpp
Closed a year ago
Random spikes of up to 30ms in ggml_cuda_op device synchronization when using a low -ngl count with dual GPU
Closed a year ago4
falcon-main.exe Exits Unexpectedly after 'Numa' Commit
Closed a year ago3
Problem with cMake on Linux focal, Cuda
Updated a year ago12
A strange delay happens after about 200 tokens
Closed a year ago4
Slowdown with tokens
Updated a year ago27
Loss of context in batch prompt processing - once again
Closed a year ago2
Windows Installation Video Tutorial
Closed a year ago1
CUDA mul_mat using cuBLAS for 3d multiplication fails on lm_head only for Falcon 7B
Closed a year ago2
Parameter --reverse-prompt won't accept text
Closed a year ago1
slow on 3090 and very high cpu usage
Updated a year ago6
Illegal instruction (core dumped) on Ubuntu 22.04.2
Updated a year ago13
mmap fails on WSL (linux too?)
Closed a year ago1
Unable to make falcon_main
Closed a year ago2
Why divert from the default GGML versions?
Updated a year ago2
falcon_main chokes on larger prompts (linux)
Closed a year ago16