e88 88e d8
d888 888b 8888 8888 ,"Y88b 888 8e d88
C8888 8888D 8888 8888 "8" 888 888 88b d88888
Y888 888P Y888 888P ,ee 888 888 888 888
"88 88" "88 88" "88 888 888 888 888
b
8b,
e88'Y88 d8 888
d888 'Y ,"Y88b 888,8, d88 ,e e, 888
C8888 "8" 888 888 " d88888 d88 88b 888
Y888 ,d ,ee 888 888 888 888 , 888
"88,d88 "88 888 888 888 "YeeP" 888
PROUDLY PRESENTS
Quantized with love from fp16 using the alpha=32 version.
Original model author: rAIfle
Original model README here and below:
Oh boy, here we go. Low-rank (r=16, alpha=32) 16bit-LoRA on top of WizardLM-2-8x22B, trained on 2 epochs of (cleaned & deduped) c2-logs. As far as I can tell, this is an upgrade from WizardLM-2-8x22B for RP purposes.
Alongside this ready-to-use release I'm also releasing the LoRA itself as well as the earlier epoch1-checkpoint of the LoRA.
The choice was fully intentional. I briefly considered a FFT but for this particular use-case a LoRA seemed a better fit. WizardLM-2-8x22B is smart by itself but its used vocabulary leaves much to be desired when it comes to RP. By training a low-rank LoRA on top of it to teach it some of Claude's writing style, we remedy that.
SorcererLM-folder.The main shoutout I want to make is to my Cartel bros, Envoid and particularly I^2, for being amazing. I count this as a team effort, so they deserve kudos too if you like this.
Trained using qlora-pipe. Configs included in the train-subfolder.
... n/a
SorcererLM-8x22b-iMat-Q2_K-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q2_K-00002-of-00002.gguf 2.3 GBSorcererLM-8x22b-iMat-IQ3_XXS-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-IQ3_XXS-00002-of-00002.gguf 5.0 GBSorcererLM-8x22b-iMat-IQ3_XS-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-IQ3_XS-00002-of-00002.gguf 8.5 GBSorcererLM-8x22b-iMat-IQ3_S-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-IQ3_S-00002-of-00002.gguf 12 GBSorcererLM-8x22b-iMat-Q3_K_S-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q3_K_S-00002-of-00002.gguf 12 GBSorcererLM-8x22b-iMat-IQ3_M-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-IQ3_M-00002-of-00002.gguf 15 GBSorcererLM-8x22b-iMat-Q3_K_M-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q3_K_M-00002-of-00002.gguf 18 GBSorcererLM-8x22b-iMat-Q3_K_L-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q3_K_L-00002-of-00002.gguf 23 GBSorcererLM-8x22b-iMat-IQ4_XS-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-IQ4_XS-00002-of-00002.gguf 26 GBSorcererLM-8x22b-iMat-Q4_K_S-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q4_K_S-00002-of-00002.gguf 31 GBSorcererLM-8x22b-iMat-Q4_K_M-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q4_K_M-00002-of-00002.gguf 36 GBSorcererLM-8x22b-iMat-Q5_K_S-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q5_K_S-00002-of-00002.gguf 47 GBSorcererLM-8x22b-iMat-Q5_K_M-00002-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q5_K_M-00001-of-00002.gguf 50 GBSorcererLM-8x22b-iMat-Q6_K-00001-of-00003.gguf 50 GBSorcererLM-8x22b-iMat-Q6_K-00002-of-00003.gguf 49 GBSorcererLM-8x22b-iMat-Q6_K-00003-of-00003.gguf 17 GBSorcererLM-8x22b-iMat-Q8_0-00001-of-00003.gguf 50 GBSorcererLM-8x22b-iMat-Q8_0-00003-of-00003.gguf 50 GBSorcererLM-8x22b-iMat-Q8_0-00002-of-00003.gguf 50 GB有文件?提交种子。
提交种子