rixty
/
KoboldAI-Client
mirror of https://github.com/KoboldAI/KoboldAI-Client.git
1
1
Fork 0
Code Issues Projects Releases Wiki Activity

Merge pull request #83 from henk717/united 

Update
This commit is contained in:
ebolam 2022-08-30 15:03:34 -04:00 committed by GitHub
parent b5a6b44582 1641b7538b
commit 8d3eb44d2e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
8 changed files with 163 additions and 61 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

118
aiserver.py
View File

@ -963,7 +963,10 @@ def loadmodelsettings():
if("nobreakmodel" in js):
vars.nobreakmodel = js["nobreakmodel"]
if("sampler_order" in js):
vars.sampler_order = js["sampler_order"]
sampler_order = vars.sampler_order
if(len(sampler_order) < 7):
sampler_order = [6] + sampler_order
vars.sampler_order = sampler_order
if("temp" in js):
vars.temp = js["temp"]
if("top_p" in js):
@ -1094,7 +1097,10 @@ def processsettings(js):
if("andepth" in js):
vars.andepth = js["andepth"]
if("sampler_order" in js):
vars.sampler_order = js["sampler_order"]
sampler_order = vars.sampler_order
if(len(sampler_order) < 7):
sampler_order = [6] + sampler_order
vars.sampler_order = sampler_order
if("temp" in js):
vars.temp = js["temp"]
if("top_p" in js):
@ -1474,22 +1480,22 @@ def get_model_info(model, directory=""):
def get_layer_count(model, directory=""):
if(model not in ["InferKit", "Colab", "API", "OAI", "GooseAI" , "ReadOnly", "TPUMeshTransformerGPTJ"]):
if(vars.model == "GPT2Custom"):
model_config = open(vars.custmodpth + "/config.json", "r")
if(model == "GPT2Custom"):
with open(os.path.join(directory, "config.json"), "r") as f:
model_config = json.load(f)
# Get the model_type from the config or assume a model type if it isn't present
else:
if(directory):
model = directory
from transformers import AutoConfig
if directory == "":
model_config = AutoConfig.from_pretrained(model, revision=vars.revision, cache_dir="cache")
if(os.path.isdir(model.replace('/', '_'))):
model_config = AutoConfig.from_pretrained(model.replace('/', '_'), revision=vars.revision, cache_dir="cache")
elif(os.path.isdir("models/{}".format(model.replace('/', '_')))):
model_config = AutoConfig.from_pretrained("models/{}".format(model.replace('/', '_')), revision=vars.revision, cache_dir="cache")
elif(os.path.isdir(directory)):
model_config = AutoConfig.from_pretrained(directory, revision=vars.revision, cache_dir="cache")
elif(os.path.isdir(vars.custmodpth.replace('/', '_'))):
model_config = AutoConfig.from_pretrained(vars.custmodpth.replace('/', '_'), revision=vars.revision, cache_dir="cache")
else:
model_config = AutoConfig.from_pretrained(vars.custmodpth, revision=vars.revision, cache_dir="cache")
model_config = AutoConfig.from_pretrained(model, revision=vars.revision, cache_dir="cache")
return utils.num_layers(model_config)
else:
return None
@ -1727,8 +1733,6 @@ def patch_transformers():
dynamic_processor_wrap(TailFreeLogitsWarper, "tfs", "tfs", cond=lambda x: x < 1.0)
dynamic_processor_wrap(TypicalLogitsWarper, "typical", "typical", cond=lambda x: x < 1.0)
dynamic_processor_wrap(TemperatureLogitsWarper, "temperature", "temp", cond=lambda x: x != 1.0)
RepetitionPenaltyLogitsProcessor.__init__ = AdvancedRepetitionPenaltyLogitsProcessor.__init__
RepetitionPenaltyLogitsProcessor.__call__ = AdvancedRepetitionPenaltyLogitsProcessor.__call__
class LuaLogitsProcessor(LogitsProcessor):
@ -1805,9 +1809,13 @@ def patch_transformers():
self.__warper_list.append(TailFreeLogitsWarper(tfs=0.5, min_tokens_to_keep=1 + (beams > 1)))
self.__warper_list.append(TypicalLogitsWarper(typical=0.5, min_tokens_to_keep=1 + (beams > 1)))
self.__warper_list.append(TemperatureLogitsWarper(temperature=0.5))
self.__warper_list.append(AdvancedRepetitionPenaltyLogitsProcessor())
def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, *args, **kwargs):
for k in vars.sampler_order:
sampler_order = vars.sampler_order[:]
if len(sampler_order) < 7: # Add repetition penalty at beginning if it's not present
sampler_order = [6] + sampler_order
for k in sampler_order:
scores = self.__warper_list[k](input_ids, scores, *args, **kwargs)
return scores
@ -1940,22 +1948,24 @@ def reset_model_settings():
vars.badwordsids = []
vars.fp32_model = False # Whether or not the most recently loaded HF model was in fp32 format
vars.modeldim = -1 # Embedding dimension of your model (e.g. it's 4096 for GPT-J-6B and 2560 for GPT-Neo-2.7B)
vars.sampler_order = [0, 1, 2, 3, 4, 5]
vars.sampler_order = [6, 0, 1, 2, 3, 4, 5]
vars.newlinemode = "n"
vars.revision = None
def load_model(use_gpu=True, gpu_layers=None, disk_layers=None, initial_load=False, online_model=""):
def load_model(use_gpu=True, gpu_layers=None, disk_layers=None, initial_load=False, online_model="", use_breakmodel_args=False, breakmodel_args_default_to_cpu=False):
global model
global generator
global torch
global model_config
global GPT2TokenizerFast
global tokenizer
if(initial_load):
use_breakmodel_args = True
reset_model_settings()
if not utils.HAS_ACCELERATE:
disk_layers = None
vars.noai = False
if not initial_load:
if not use_breakmodel_args:
set_aibusy(True)
if vars.model != 'ReadOnly':
emit('from_server', {'cmd': 'model_load_status', 'data': "Loading {}".format(vars.model)}, broadcast=True)
@ -1963,12 +1973,16 @@ def load_model(use_gpu=True, gpu_layers=None, disk_layers=None, initial_load=Fal
time.sleep(0.1)
if gpu_layers is not None:
args.breakmodel_gpulayers = gpu_layers
elif initial_load:
elif use_breakmodel_args:
gpu_layers = args.breakmodel_gpulayers
if breakmodel_args_default_to_cpu and gpu_layers is None:
gpu_layers = args.breakmodel_gpulayers = []
if disk_layers is not None:
args.breakmodel_disklayers = int(disk_layers)
elif initial_load:
elif use_breakmodel_args:
disk_layers = args.breakmodel_disklayers
if breakmodel_args_default_to_cpu and disk_layers is None:
disk_layers = args.breakmodel_disklayers = 0
#We need to wipe out the existing model and refresh the cuda cache
model = None
@ -2062,6 +2076,7 @@ def load_model(use_gpu=True, gpu_layers=None, disk_layers=None, initial_load=Fal
if(not vars.use_colab_tpu and vars.model not in ["InferKit", "Colab", "API", "OAI", "GooseAI" , "ReadOnly", "TPUMeshTransformerGPTJ", "TPUMeshTransformerGPTNeoX"]):
loadmodelsettings()
loadsettings()
print(2)
print("{0}Looking for GPU support...{1}".format(colors.PURPLE, colors.END), end="")
vars.hascuda = torch.cuda.is_available()
vars.bmsupported = (utils.HAS_ACCELERATE or vars.model_type in ("gpt_neo", "gptj", "xglm", "opt")) and not vars.nobreakmodel
@ -2311,7 +2326,6 @@ def load_model(use_gpu=True, gpu_layers=None, disk_layers=None, initial_load=Fal
# If we're using torch_lazy_loader, we need to get breakmodel config
# early so that it knows where to load the individual model tensors
if (utils.HAS_ACCELERATE or vars.lazy_load and vars.hascuda and vars.breakmodel) and not vars.nobreakmodel:
print(1)
device_config(model_config)
# Download model from Huggingface if it does not exist, otherwise load locally
@ -2551,8 +2565,11 @@ def load_model(use_gpu=True, gpu_layers=None, disk_layers=None, initial_load=Fal
vars.compiling = False
def tpumtjgenerate_settings_callback() -> dict:
sampler_order = vars.sampler_order[:]
if len(sampler_order) < 7: # Add repetition penalty at beginning if it's not present
sampler_order = [6] + sampler_order
return {
"sampler_order": vars.sampler_order,
"sampler_order": sampler_order,
"top_p": float(vars.top_p),
"temp": float(vars.temp),
"top_k": int(vars.top_k),
@ -3659,12 +3676,16 @@ def get_message(msg):
sendUSStatItems()
elif(msg['cmd'] == 'samplers'):
sampler_order = msg["data"]
sampler_order_min_length = 6
sampler_order_max_length = 7
if(not isinstance(sampler_order, list)):
raise ValueError(f"Sampler order must be a list, but got a {type(sampler_order)}")
if(len(sampler_order) != len(vars.sampler_order)):
raise ValueError(f"Sampler order must be a list of length {len(vars.sampler_order)}, but got a list of length {len(sampler_order)}")
if(not (sampler_order_min_length <= len(sampler_order) <= sampler_order_max_length)):
raise ValueError(f"Sampler order must be a list of length greater than or equal to {sampler_order_min_length} and less than or equal to {sampler_order_max_length}, but got a list of length {len(sampler_order)}")
if(not all(isinstance(e, int) for e in sampler_order)):
raise ValueError(f"Sampler order must be a list of ints, but got a list with at least one non-int element")
if(min(sampler_order) != 0 or max(sampler_order) != len(sampler_order) - 1 or len(set(sampler_order)) != len(sampler_order)):
raise ValueError(f"Sampler order list of length {len(sampler_order)} must be a permutation of the first {len(sampler_order)} nonnegative integers")
vars.sampler_order = sampler_order
settingschanged()
elif(msg['cmd'] == 'list_model'):
@ -3848,7 +3869,6 @@ def get_message(msg):
emit(
'from_server',
{'cmd': 'showfieldbudget', 'data': {"length": None, "max": None, "field": field}},
broadcast=True
)
return
@ -4618,7 +4638,7 @@ def _generate(txt, minimum, maximum, found_entries):
gen_in,
do_sample=True,
max_length=int(2e9),
repetition_penalty=1.1,
repetition_penalty=1.0,
bad_words_ids=vars.badwordsids,
use_cache=True,
num_return_sequences=numseqs
@ -7242,6 +7262,9 @@ class WorldInfoFoldersUIDsSchema(KoboldSchema):
class WorldInfoUIDsSchema(WorldInfoEntriesUIDsSchema):
folders: List[WorldInfoFolderSchema] = fields.List(fields.Nested(WorldInfoFolderUIDsSchema), required=True)
class ModelSelectionSchema(KoboldSchema):
model: str = fields.String(required=True, validate=validate.Regexp(r"^(?!\s*NeoCustom)(?!\s*GPT2Custom)(?!\s*TPUMeshTransformerGPTJ)(?!\s*TPUMeshTransformerGPTNeoX)(?!\s*GooseAI)(?!\s*OAI)(?!\s*InferKit)(?!\s*Colab)(?!\s*API).*$"), metadata={"description": 'Hugging Face model ID, the path to a model folder (relative to the "models" folder in the KoboldAI root folder) or "ReadOnly" for no model'})
def _generate_text(body: GenerationInputSchema):
if vars.aibusy or vars.genseqs:
abort(Response(json.dumps({"detail": {
@ -7453,6 +7476,49 @@ def get_model():
return {"result": vars.model}
@api_v1.put("/model")
@api_schema_wrap
def put_model(body: ModelSelectionSchema):
"""---
put:
summary: Load a model
description: |-2
Loads a model given its Hugging Face model ID, the path to a model folder (relative to the "models" folder in the KoboldAI root folder) or "ReadOnly" for no model.
tags:
- model
requestBody:
required: true
content:
application/json:
schema: ModelSelectionSchema
example:
model: ReadOnly
responses:
200:
description: Successful request
content:
application/json:
schema: EmptySchema
{api_validation_error_response}
{api_server_busy_response}
"""
if vars.aibusy or vars.genseqs:
abort(Response(json.dumps({"detail": {
"msg": "Server is busy; please try again later.",
"type": "service_unavailable",
}}), mimetype="application/json", status=503))
set_aibusy(1)
old_model = vars.model
vars.model = body.model.strip()
try:
load_model(use_breakmodel_args=True, breakmodel_args_default_to_cpu=True)
except Exception as e:
vars.model = old_model
raise e
set_aibusy(0)
return {}
def prompt_validator(prompt: str):
if len(prompt.strip()) == 0:
raise ValidationError("String does not match expected pattern.")

10
colab/TPU.ipynb
View File

@ -66,7 +66,7 @@
"#@title <b><-- Select your model below and then click this to start KoboldAI</b>\n",
"#@markdown You can find a description of the models below along with instructions on how to start KoboldAI.\n",
"\n",
"Model = \"Nerys 13B V2\" #@param [\"Nerys 13B V2\", \"Janeway 13B\", \"Shinen 13B\", \"Skein 6B\", \"Janeway 6B\", \"Adventure 6B\", \"Shinen 6B\", \"Lit 6B\", \"NeoX 20B\", \"OPT 13B\", \"Fairseq Dense 13B\", \"GPT-J-6B\"] {allow-input: true}\n",
"Model = \"Nerys 13B V2\" #@param [\"Nerys 13B V2\", \"Janeway 13B\", \"Shinen 13B\", \"Skein 20B\", \"Skein 6B\", \"Janeway 6B\", \"Adventure 6B\", \"Shinen 6B\", \"Lit 6B\", \"NeoX 20B\", \"OPT 13B\", \"Fairseq Dense 13B\", \"GPT-J-6B\"] {allow-input: true}\n",
"Version = \"Official\" #@param [\"Official\", \"United\"] {allow-input: true}\n",
"Provider = \"Cloudflare\" #@param [\"Localtunnel\", \"Cloudflare\"]\n",
"\n",
@ -93,6 +93,10 @@
" Model = \"KoboldAI/fairseq-dense-13B-Shinen\"\n",
" path = \"\"\n",
" download = \"\"\n",
"elif Model == \"Skein 20B\":\n",
" Model = \"KoboldAI/GPT-NeoX-20B-Skein\"\n",
" path = \"\"\n",
" download = \"\"\n",
"elif Model == \"NeoX 20B\":\n",
" Model = \"EleutherAI/gpt-neox-20b\"\n",
" path = \"\"\n",
@ -128,7 +132,7 @@
"elif Model == \"GPT-J-6B\":\n",
" Model = \"EleutherAI/gpt-j-6B\"\n",
" path = \"\"\n",
" download = \"\"\n",
" download = \"\"\n",
"else:\n",
" path = \"\"\n",
" download = \"\"\n",
@ -225,4 +229,4 @@
},
"nbformat": 4,
"nbformat_minor": 0
}
}

11
gensettings.py
View File

@ -274,6 +274,17 @@ gensettingstf = [
"default": 0,
"tooltip": "Shows token selection probabilities. Does not work with more than one gens per action."
},
{
"uitype": "toggle",
"unit": "bool",
"label": "Show Field Budget",
"id": "setshowbudget",
"min": 0,
"max": 1,
"step": 1,
"default": 0,
"tooltip": "Shows token usage when typing in relevant text boxes. <b>May lag slower devices.</b>"
},
]
gensettingsik =[{

27
static/application.js
View File

@ -241,8 +241,27 @@ function addSetting(ob) {
if(ob.id == "setadventure"){
setadventure($(this).prop('checked'));
}
});
}
if (ob.id === "setshowbudget") {
$("#setshowbudget").on("change", function () {
for (const el of document.getElementsByClassName("input-token-usage")) {
if (this.checked) {
el.classList.remove("hidden");
} else {
el.classList.add("hidden");
}
}
});
if (!$("#setshowbudget")[0].checked) {
for (const el of document.getElementsByClassName("input-token-usage")) {
el.classList.add("hidden");
}
}
}
}
function refreshTitle() {
@ -1287,12 +1306,13 @@ function buildSamplerList(samplers) {
"Tail-free Sampling",
"Typical Sampling",
"Temperature",
"Repetition Penalty",
]
for(i=0; i<samplers.length; i++) {
samplerslist.append("<div class=\"flex\">\
<div class=\"samplerslistitem flex-row-container\" sid=\""+samplers[i]+"\">\
<div class=\"flex-row\">\
<div>"+samplers_lookup_table[samplers[i]]+"</div>\
<div>"+(samplers[i] < samplers_lookup_table.length ? samplers_lookup_table[samplers[i]] : "Unknown sampler #" + samplers[i])+"</div>\
</div>\
</div>\
</div>");
@ -2165,6 +2185,9 @@ function interpolateRGB(color0, color1, t) {
}
function updateInputBudget(inputElement) {
let budgetElement = document.getElementById("setshowbudget");
if (budgetElement && !budgetElement.checked) return;
let data = {"unencoded": inputElement.value, "field": inputElement.id};
if (inputElement.id === "anoteinput") {
@ -2182,7 +2205,6 @@ function registerTokenCounters() {
let span = document.createElement("span");
span.classList.add("input-token-usage");
span.innerText = "?/? Tokens";
el.appendChild(span);
let inputElement = el.querySelector("input, textarea");
@ -2958,6 +2980,7 @@ $(document).ready(function(){
$("#showmodelnamecontainer").removeClass("hidden");
} else if(msg.cmd == 'hide_model_name') {
$("#showmodelnamecontainer").addClass("hidden");
$(window).off('beforeunload');
location.reload();
//console.log("Closing window");
} else if(msg.cmd == 'model_load_status') {

2
static/custom.css
View File

@ -473,7 +473,7 @@ body.connected #popupfooter, #popupfooter.always-available {
}
#samplerslist {
height: 300px;
height: 310px;
overflow-y: scroll;
overflow-wrap: anywhere;
}

48
tpu_mtj_backend.py
View File

@ -176,7 +176,7 @@ def apply_repetition_penalty_dynamic(logits, tokens, repetition_penalty, generat
logits[tokens] = penalty_logits
return logits
def kobold_sample_dynamic(key, logits, sampler_order: Optional[np.ndarray] = None, top_p=0.9, temp=0.5, top_k=0, tfs=1.0, typical=1.0, top_a=0.0):
def kobold_sample_dynamic(key, logits, rpargs, sampler_order: Optional[np.ndarray] = None, top_p=0.9, temp=0.5, top_k=0, tfs=1.0, typical=1.0, top_a=0.0):
'''
This gets called by generate_loop_fn to apply a series of 6 filters
to the logits (top-k, then top-a, then top-p, then TFS, then typical, then temperature)
@ -312,6 +312,7 @@ def kobold_sample_dynamic(key, logits, sampler_order: Optional[np.ndarray] = Non
if k == 3 and tfs < 1.0: logits = tail_free_filter(logits)
if k == 4 and typical < 1.0: logits = typical_filter(logits)
if k == 5 and temp != 1.0: logits = temp_filter(logits)
if k == 6 and rpargs[1] != 1.0: logits = apply_repetition_penalty_dynamic(logits, *rpargs)
# Finally, pick one token using the softmax thingy again (it gives
# an array whose elements sum to 1 so it can be used nicely as a
# probability distribution)
@ -362,7 +363,7 @@ def apply_repetition_penalty_static(logits, tokens, repetition_penalty, generate
# positions in the logits array
return logits.at[tokens].set(penalty_logits)
def kobold_sample_static(key, logits, sampler_order: Optional[np.ndarray] = None, top_p=0.9, temp=0.5, top_k=0, tfs=1.0, typical=1.0, top_a=0.0):
def kobold_sample_static(key, logits, rpargs, sampler_order: Optional[np.ndarray] = None, top_p=0.9, temp=0.5, top_k=0, tfs=1.0, typical=1.0, top_a=0.0):
'''
This gets called by generate_loop_fn to apply a series of 6 filters
to the logits (top-k, then top-a, then top-p, then TFS, then typical, then temperature)
@ -497,6 +498,7 @@ def kobold_sample_static(key, logits, sampler_order: Optional[np.ndarray] = None
logits = jax.lax.cond(jnp.logical_and(k == 3, tfs < 1.0), tail_free_filter, lambda x: x, logits)
logits = jax.lax.cond(jnp.logical_and(k == 4, typical < 1.0), typical_filter, lambda x: x, logits)
logits = jax.lax.cond(jnp.logical_and(k == 5, temp != 1.0), temp_filter, lambda x: x, logits)
logits = jax.lax.cond(jnp.logical_and(k == 6, rpargs[1] != 1.0), lambda x: apply_repetition_penalty_static(*x), lambda x: x[0], (logits, *rpargs))
# Finally, pick one token using the softmax thingy again (it gives
# an array whose elements sum to 1 so it can be used nicely as a
# probability distribution)
@ -513,17 +515,6 @@ def sample_func(data, key, numseqs_aux, badwords, repetition_penalty, generated_
# Get the pseudo-random number generator key that will
# be used by kobold_sample_dynamic to randomly pick a token
sample_key, new_key = jax.random.split(sample_key, num=2)
# Apply repetition penalty to all tokens that are
# currently inside the "generated" array
logits = apply_repetition_penalty_dynamic(
logits,
generated,
repetition_penalty,
generated_index,
gen_length,
rpslope,
rprange,
)
# Remove any tokens in the badwords list by setting
# their logits to negative infinity which effectively
# makes their probabilities of being chosen zero
@ -535,6 +526,14 @@ def sample_func(data, key, numseqs_aux, badwords, repetition_penalty, generated_
next_token = kobold_sample_dynamic(
sample_key,
logits,
(
generated,
repetition_penalty,
generated_index,
gen_length,
rpslope,
rprange,
),
**sampler_options,
)
# Remember what token was picked
@ -606,18 +605,6 @@ class PenalizingCausalTransformer(CausalTransformer):
assert logits.shape == (1, config["n_vocab"])
# Flatten it into a 1D array to make it easier to use
logits = logits[0]
# Apply repetition penalty to all tokens that are
# currently inside the "generated" array
if repetition_penalty is not None:
logits = apply_repetition_penalty_static(
logits,
generated,
repetition_penalty,
generated_index,
gen_length,
rpslope,
rprange,
)
# Remove any tokens in the badwords list by setting
# their logits to negative infinity which effectively
# makes their probabilities of being chosen zero
@ -629,6 +616,14 @@ class PenalizingCausalTransformer(CausalTransformer):
next_token = kobold_sample_static(
sample_key,
logits,
(
generated,
repetition_penalty,
generated_index,
gen_length,
rpslope,
rprange,
),
**sampler_options,
)
# Remember what token was picked
@ -863,6 +858,9 @@ def infer_static(
maps.thread_resources.env = thread_resources_env
if sampler_order is None:
sampler_order = utils.default_sampler_order.copy()
sampler_order = sampler_order[:]
if len(sampler_order) < 7: # Add repetition penalty at beginning if it's not present
sampler_order = [6] + sampler_order
sampler_order = np.uint32(sampler_order)
total_batch = 1
tokens = context

4
utils.py
View File

@ -33,7 +33,7 @@ layers_module_names: Optional[List[str]] = None
module_names: Optional[List[str]] = None
named_buffers: Optional[List[tuple]] = None
default_sampler_order = [0, 1, 2, 3, 4, 5]
default_sampler_order = [6, 0, 1, 2, 3, 4, 5]
#==================================================================#
# Decorator to prevent a function's actions from being run until
@ -167,7 +167,7 @@ def decodenewlines(txt):
# Returns number of layers given an HF model config
#==================================================================#
def num_layers(config):
return config.num_layers if hasattr(config, "num_layers") else config.n_layer if hasattr(config, "n_layer") else config.num_hidden_layers if hasattr(config, 'num_hidden_layers') else None
return config["n_layer"] if isinstance(config, dict) else config.num_layers if hasattr(config, "num_layers") else config.n_layer if hasattr(config, "n_layer") else config.num_hidden_layers if hasattr(config, 'num_hidden_layers') else None
#==================================================================#
# Downloads huggingface checkpoints using aria2c if possible

4
warpers.py
View File

@ -28,10 +28,10 @@ SOFTWARE.
'''
import torch
from transformers import LogitsWarper, LogitsProcessor
from transformers import LogitsWarper
class AdvancedRepetitionPenaltyLogitsProcessor(LogitsProcessor):
class AdvancedRepetitionPenaltyLogitsProcessor(LogitsWarper):
def __init__(self, *args, **kwargs):
pass