BOS is actually the EOS token by default
#30 opened about 1 year ago
by
dblakely
Adding Evaluation Results
#29 opened about 1 year ago
by
leaderboard-pr-bot
Adding `safetensors` variant of this model
#28 opened over 1 year ago
by
qeternity
set use_cache=true for faster decoding
#27 opened over 1 year ago
by
zxcvvxcz
Set model_max_length of tokenizer equal to the max_length of the model
#26 opened over 1 year ago
by
zxcvvxcz
Something's Wrong, It's Not Your Fault, Llama 2 Is Fundamentally Flawed
1
#25 opened over 1 year ago
by
deleted
Update README.md
#16 opened over 1 year ago
by
haipeng1
mat1 and mat2 shapes cannot be multiplied (38x5120 and 1x2560)
1
#13 opened over 1 year ago
by
LaferriereJC
Request: Wizardlm-22b
#12 opened over 1 year ago
by
rombodawg
inference speed is considerably slow
#11 opened over 1 year ago
by
sonald
Should the "Max Positional Embeddings" be 4096?
1
#10 opened over 1 year ago
by
barleyspectacular
Feel free to use my datasets for further refinement!
#9 opened over 1 year ago
by
rombodawg
Missing model card & datasets info
#8 opened over 1 year ago
by
markding
RuntimeError: probability tensor contains either `inf`, `nan` or element < 0
#7 opened over 1 year ago
by
Satya93
Still non-commercial?
#6 opened over 1 year ago
by
kalijason
RuntimeError: expected scalar type Half but found Char
2
#5 opened over 1 year ago
by
nicolasbo
database connection ?
#4 opened over 1 year ago
by
nobitha
What is the prompt format?
1
#3 opened over 1 year ago
by
TheBloke
What is this model based off of?
1
#2 opened over 1 year ago
by
rombodawg
Dataset Availability?
#1 opened over 1 year ago
by
jonfairbanks