2023-05-19 16:42:41 +02:00
|
|
|
.*(llama|alpac|vicuna|guanaco|koala|llava|wizardlm|metharme|pygmalion-7b|wizard-mega|openbuddy|vigogne|h2ogpt-research|manticore):
|
2023-04-14 16:07:28 +02:00
|
|
|
model_type: 'llama'
|
2023-05-10 07:07:22 +02:00
|
|
|
.*(opt-|opt_|opt1|opt3|optfor|galactica|galpaca|pygmalion-350m):
|
|
|
|
model_type: 'opt'
|
2023-05-19 16:42:41 +02:00
|
|
|
.*(gpt-j|gptj|gpt4all-j|malion-6b|pygway|pygmalion-6b|dolly-v1):
|
2023-05-10 07:07:22 +02:00
|
|
|
model_type: 'gptj'
|
2023-05-19 16:42:41 +02:00
|
|
|
.*(gpt-neox|koalpaca-polyglot|polyglot.*koalpaca|polyglot-ko|polyglot_ko|pythia|stablelm|incite|dolly-v2|polycoder|h2ogpt-oig|h2ogpt-oasst1|h2ogpt-gm):
|
2023-08-14 20:17:24 +02:00
|
|
|
model_type: 'gptneox'
|
2023-05-19 16:42:41 +02:00
|
|
|
.*llama:
|
|
|
|
model_type: 'llama'
|
|
|
|
.*bloom:
|
|
|
|
model_type: 'bloom'
|
2023-08-14 04:09:03 +02:00
|
|
|
.*gpt2:
|
|
|
|
model_type: 'gpt2'
|
|
|
|
.*falcon:
|
|
|
|
model_type: 'falcon'
|
|
|
|
.*mpt:
|
|
|
|
model_type: 'mpt'
|
|
|
|
.*(starcoder|starchat):
|
2023-08-14 20:17:24 +02:00
|
|
|
model_type: 'starcoder'
|
2023-08-14 04:09:03 +02:00
|
|
|
.*dolly-v2:
|
2023-08-14 20:17:24 +02:00
|
|
|
model_type: 'dollyv2'
|
2023-08-14 04:09:03 +02:00
|
|
|
.*replit:
|
|
|
|
model_type: 'replit'
|
2023-05-19 16:42:41 +02:00
|
|
|
llama-65b-gptq-3bit:
|
|
|
|
groupsize: 'None'
|
2023-05-10 01:37:31 +02:00
|
|
|
.*(4bit|int4):
|
2023-04-14 16:07:28 +02:00
|
|
|
wbits: 4
|
2023-05-10 01:37:31 +02:00
|
|
|
.*(3bit|int3):
|
2023-04-14 16:07:28 +02:00
|
|
|
wbits: 3
|
2023-05-10 01:37:31 +02:00
|
|
|
.*(-2bit|_2bit|int2-):
|
|
|
|
wbits: 2
|
|
|
|
.*(-1bit|_1bit|int1-):
|
|
|
|
wbits: 1
|
|
|
|
.*(8bit|int8):
|
|
|
|
wbits: 8
|
|
|
|
.*(-7bit|_7bit|int7-):
|
|
|
|
wbits: 7
|
|
|
|
.*(-6bit|_6bit|int6-):
|
|
|
|
wbits: 6
|
|
|
|
.*(-5bit|_5bit|int5-):
|
|
|
|
wbits: 5
|
2023-05-12 11:09:45 +02:00
|
|
|
.*(-gr32-|-32g-|groupsize32|-32g$):
|
2023-05-10 01:37:31 +02:00
|
|
|
groupsize: 32
|
2023-05-12 11:11:58 +02:00
|
|
|
.*(-gr64-|-64g-|groupsize64|-64g$):
|
2023-05-10 01:37:31 +02:00
|
|
|
groupsize: 64
|
|
|
|
.*(gr128|128g|groupsize128):
|
2023-04-14 16:07:28 +02:00
|
|
|
groupsize: 128
|
2023-05-10 01:37:31 +02:00
|
|
|
.*(gr1024|1024g|groupsize1024):
|
|
|
|
groupsize: 1024
|
2023-05-19 16:42:41 +02:00
|
|
|
.*(oasst|openassistant-|stablelm-7b-sft-v7-epoch-3):
|
2023-04-14 16:07:28 +02:00
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Open Assistant'
|
2023-05-10 01:37:31 +02:00
|
|
|
skip_special_tokens: false
|
2023-05-19 16:42:41 +02:00
|
|
|
(?!.*galactica)(?!.*reward).*openassistant:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Open Assistant'
|
|
|
|
skip_special_tokens: false
|
|
|
|
(?!.*v0)(?!.*1.1)(?!.*1_1)(?!.*stable)(?!.*chinese).*vicuna:
|
2023-04-14 16:07:28 +02:00
|
|
|
mode: 'instruct'
|
2023-04-26 21:20:27 +02:00
|
|
|
instruction_template: 'Vicuna-v0'
|
2023-05-10 01:37:31 +02:00
|
|
|
.*vicuna.*v0:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vicuna-v0'
|
2023-06-21 17:30:44 +02:00
|
|
|
.*vicuna.*(1.1|1_1|1.3|1_3):
|
2023-05-10 01:37:31 +02:00
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vicuna-v1.1'
|
2023-08-10 18:42:24 +02:00
|
|
|
.*vicuna.*(1.5|1_5):
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vicuna-v1.1'
|
|
|
|
truncation_length: 4096
|
|
|
|
rms_norm_eps: 5.0e-6
|
2023-05-10 01:37:31 +02:00
|
|
|
.*stable.*vicuna:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'StableVicuna'
|
2023-05-19 16:42:41 +02:00
|
|
|
(?!.*chat).*chinese-vicuna:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*chinese-vicuna.*chat:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Chinese-Vicuna-Chat'
|
2023-04-14 16:07:28 +02:00
|
|
|
.*alpaca:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
2023-04-14 16:15:59 +02:00
|
|
|
.*alpaca-native-4bit:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
wbits: 4
|
|
|
|
groupsize: 128
|
2023-05-10 01:37:31 +02:00
|
|
|
.*galactica:
|
2023-04-16 19:24:49 +02:00
|
|
|
skip_special_tokens: false
|
|
|
|
.*dolly-v[0-9]-[0-9]*b:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
skip_special_tokens: false
|
|
|
|
custom_stopping_strings: '"### End"'
|
2023-04-16 19:40:45 +02:00
|
|
|
.*koala:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Koala'
|
2023-04-17 00:15:03 +02:00
|
|
|
.*chatglm:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'ChatGLM'
|
2023-05-10 03:29:22 +02:00
|
|
|
.*metharme:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Metharme'
|
2023-04-24 01:32:22 +02:00
|
|
|
.*llava:
|
|
|
|
mode: 'instruct'
|
|
|
|
model_type: 'llama'
|
|
|
|
instruction_template: 'LLaVA'
|
2023-04-26 08:47:34 +02:00
|
|
|
custom_stopping_strings: '"\n###"'
|
2023-04-26 08:21:53 +02:00
|
|
|
.*raven:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'RWKV-Raven'
|
models/config.yaml: +platypus/gplatty, +longchat, +vicuna-33b, +Redmond-Hermes-Coder, +wizardcoder, +more (#2928)
* +platypus/gplatty
* +longchat, +vicuna-33b, +Redmond-Hermes-Coder
* +wizardcoder
* +superplatty
* +Godzilla, +WizardLM-V1.1, +rwkv 8k,
+wizard-mega fix </s>
---------
Co-authored-by: Matthew Ashton <mashton-gitlab@zhero.org>
2023-07-11 23:53:48 +02:00
|
|
|
.*ctx8192:
|
|
|
|
truncation_length: 8192
|
2023-05-05 04:19:23 +02:00
|
|
|
.*moss-moon.*sft:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'MOSS'
|
2023-05-10 01:37:31 +02:00
|
|
|
.*stablelm-tuned:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'StableLM'
|
|
|
|
truncation_length: 4096
|
|
|
|
.*stablelm-base:
|
|
|
|
truncation_length: 4096
|
|
|
|
.*galactica.*finetuned:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Galactica Finetuned'
|
|
|
|
.*galactica.*-v2:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Galactica v2'
|
|
|
|
(?!.*finetuned)(?!.*-v2).*galactica:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Galactica'
|
|
|
|
.*guanaco:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Guanaco non-chat'
|
|
|
|
.*baize:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Baize'
|
|
|
|
.*mpt-.*instruct:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*mpt-.*chat:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'MPT-Chat'
|
|
|
|
(?!.*-flan-)(?!.*-t5-).*lamini-:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*incite.*chat:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'INCITE-Chat'
|
|
|
|
.*incite.*instruct:
|
|
|
|
mode: 'instruct'
|
2023-05-11 16:06:39 +02:00
|
|
|
instruction_template: 'INCITE-Instruct'
|
2023-05-19 16:42:41 +02:00
|
|
|
.*wizard.*mega:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Wizard-Mega'
|
models/config.yaml: +platypus/gplatty, +longchat, +vicuna-33b, +Redmond-Hermes-Coder, +wizardcoder, +more (#2928)
* +platypus/gplatty
* +longchat, +vicuna-33b, +Redmond-Hermes-Coder
* +wizardcoder
* +superplatty
* +Godzilla, +WizardLM-V1.1, +rwkv 8k,
+wizard-mega fix </s>
---------
Co-authored-by: Matthew Ashton <mashton-gitlab@zhero.org>
2023-07-11 23:53:48 +02:00
|
|
|
custom_stopping_strings: '"</s>"'
|
2023-05-19 16:42:41 +02:00
|
|
|
.*ziya-:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Ziya'
|
|
|
|
.*koalpaca:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'KoAlpaca'
|
|
|
|
.*openbuddy:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'OpenBuddy'
|
|
|
|
(?!.*chat).*vigogne:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vigogne-Instruct'
|
|
|
|
.*vigogne.*chat:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vigogne-Chat'
|
|
|
|
.*(llama-deus|supercot|llama-natural-instructions|open-llama-0.3t-7b-instruct-dolly-hhrlhf|open-llama-0.3t-7b-open-instruct):
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*bactrian:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Bactrian'
|
|
|
|
.*(h2ogpt-oig-|h2ogpt-oasst1-|h2ogpt-research-oasst1-):
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'H2O-human_bot'
|
|
|
|
.*h2ogpt-gm-:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'H2O-prompt_answer'
|
|
|
|
.*manticore:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Manticore Chat'
|
2023-06-01 19:42:12 +02:00
|
|
|
.*bluemoonrp-(30|13)b:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Bluemoon'
|
2023-06-05 17:07:52 +02:00
|
|
|
truncation_length: 4096
|
2023-06-05 16:52:13 +02:00
|
|
|
.*Nous-Hermes-13b:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
2023-06-21 17:30:44 +02:00
|
|
|
.*airoboros:
|
2023-06-05 16:52:13 +02:00
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vicuna-v1.1'
|
2023-07-17 16:27:42 +02:00
|
|
|
.*airoboros.*1.2:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Airoboros-v1.2'
|
2023-06-21 17:30:44 +02:00
|
|
|
.*alpa(cino|sta):
|
2023-06-18 00:14:25 +02:00
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*hippogriff:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Hippogriff'
|
2023-06-21 17:30:44 +02:00
|
|
|
.*lazarus:
|
2023-06-18 00:14:25 +02:00
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
2023-06-21 17:30:44 +02:00
|
|
|
.*guanaco-.*(7|13|33|65)b:
|
2023-06-18 00:14:25 +02:00
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Guanaco'
|
2023-06-21 17:30:44 +02:00
|
|
|
.*hypermantis:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*open-llama-.*-open-instruct:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*starcoder-gpteacher-code-instruct:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*tulu:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Tulu'
|
|
|
|
.*chronos:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*samantha:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Samantha'
|
|
|
|
.*wizardcoder:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*starchat-beta:
|
|
|
|
mode: 'instruct'
|
2023-06-22 02:05:08 +02:00
|
|
|
instruction_template: 'Starchat-Beta'
|
2023-08-09 14:26:12 +02:00
|
|
|
custom_stopping_strings: '"<|end|>"'
|
2023-06-22 02:05:08 +02:00
|
|
|
.*minotaur:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Minotaur'
|
2023-06-24 15:14:19 +02:00
|
|
|
.*minotaur-15b:
|
2023-06-22 02:05:08 +02:00
|
|
|
truncation_length: 8192
|
2023-06-25 06:54:53 +02:00
|
|
|
.*orca_mini:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Orca Mini'
|
2023-06-28 03:05:52 +02:00
|
|
|
.*landmark:
|
|
|
|
truncation_length: 8192
|
|
|
|
.*superhot-8k:
|
|
|
|
truncation_length: 8192
|
2023-07-05 03:50:01 +02:00
|
|
|
.*xgen.*-inst:
|
2023-06-29 08:32:53 +02:00
|
|
|
truncation_length: 8192
|
2023-07-05 03:50:01 +02:00
|
|
|
instruction_template: 'Vicuna-v0'
|
models/config.yaml: +platypus/gplatty, +longchat, +vicuna-33b, +Redmond-Hermes-Coder, +wizardcoder, +more (#2928)
* +platypus/gplatty
* +longchat, +vicuna-33b, +Redmond-Hermes-Coder
* +wizardcoder
* +superplatty
* +Godzilla, +WizardLM-V1.1, +rwkv 8k,
+wizard-mega fix </s>
---------
Co-authored-by: Matthew Ashton <mashton-gitlab@zhero.org>
2023-07-11 23:53:48 +02:00
|
|
|
.*(platypus|gplatty|superplatty):
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
.*longchat:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vicuna-v1.1'
|
|
|
|
.*vicuna-33b:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vicuna-v1.1'
|
|
|
|
.*redmond-hermes-coder:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
truncation_length: 8192
|
|
|
|
.*wizardcoder-15b:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
|
|
|
truncation_length: 8192
|
2023-08-09 19:38:35 +02:00
|
|
|
.*wizardlm:
|
models/config.yaml: +platypus/gplatty, +longchat, +vicuna-33b, +Redmond-Hermes-Coder, +wizardcoder, +more (#2928)
* +platypus/gplatty
* +longchat, +vicuna-33b, +Redmond-Hermes-Coder
* +wizardcoder
* +superplatty
* +Godzilla, +WizardLM-V1.1, +rwkv 8k,
+wizard-mega fix </s>
---------
Co-authored-by: Matthew Ashton <mashton-gitlab@zhero.org>
2023-07-11 23:53:48 +02:00
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Vicuna-v1.1'
|
|
|
|
.*godzilla:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Alpaca'
|
2023-07-18 22:19:18 +02:00
|
|
|
.*llama-(2|v2):
|
|
|
|
truncation_length: 4096
|
2023-07-26 17:23:24 +02:00
|
|
|
rms_norm_eps: 5.0e-6
|
2023-07-18 22:19:18 +02:00
|
|
|
.*llama-(2|v2).*chat:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'Llama-v2'
|
2023-08-09 19:38:35 +02:00
|
|
|
.*70b.*ggml.*\.bin:
|
2023-07-25 00:26:29 +02:00
|
|
|
n_gqa: 8
|
2023-08-01 22:27:29 +02:00
|
|
|
.*newhope:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'NewHope'
|
2023-08-03 21:02:54 +02:00
|
|
|
.*stablebeluga2:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'StableBeluga2'
|
|
|
|
truncation_length: 4096
|
|
|
|
rms_norm_eps: 5.0e-6
|
2023-08-08 23:10:04 +02:00
|
|
|
.*openchat:
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'OpenChat'
|
2023-08-14 04:09:03 +02:00
|
|
|
.*falcon.*-instruct:
|
|
|
|
mode: 'instruct'
|
2023-08-18 00:03:08 +02:00
|
|
|
.*(openorca-platypus2):
|
|
|
|
mode: 'instruct'
|
|
|
|
instruction_template: 'OpenOrca-Platypus2'
|
|
|
|
custom_stopping_strings: '"### Instruction:", "### Response:"'
|
|
|
|
rms_norm_eps: 5.0e-6
|
2023-08-25 16:06:57 +02:00
|
|
|
.*codellama:
|
|
|
|
rope_freq_base: 1000000
|
|
|
|
.*codellama.*instruct:
|
|
|
|
instruction_template: 'Llama-v2'
|