iandennismiller commited on
Commit
2b59313
1 Parent(s): 0c687a2

koboldcpp launch script

Browse files

update llama-cpp-python and koboldcpp
llama.sh supports min_p sampling
llama.sh supports capybara

Files changed (4) hide show
  1. bin/llama-koboldcpp.sh +39 -0
  2. bin/llama-update.sh +22 -1
  3. bin/llama.sh +10 -4
  4. install.sh +1 -0
bin/llama-koboldcpp.sh ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+
3
+ # if conf does not exist, create it
4
+ if [ ! -f "$HOME/.config/llama/llama-koboldcpp.conf" ]; then
5
+ mkdir -p "$HOME/.config/llama"
6
+ cat <<EOF > "$HOME/.config/llama/llama-koboldcpp.conf"
7
+ LLAMA_MODEL_NAME=$HOME/.ai/models/llama/teknium/OpenHermes-2.5-Mistral-7B/openhermes-2.5-mistral-7b-f16.gguf
8
+ LLAMA_CONTEXT_SIZE=8192
9
+
10
+ KOBOLD_PORT=5000
11
+ KOBOLD_LOG=$HOME/.local/var/llama-koboldcpp.log
12
+ PYTHON_EXEC=$HOME/.virtualenvs/koboldcpp/bin/python
13
+ EOF
14
+ fi
15
+
16
+ source "$HOME/.config/llama/llama-koboldcpp.conf"
17
+
18
+ # if arg1 is "stop" then pkill the koboldcpp.py server
19
+ if [[ $# -eq 1 ]] && [[ $1 == "stop" ]]; then
20
+ echo "Stopping koboldcpp server..."
21
+ pkill -f "koboldcpp.py"
22
+ echo "ok"
23
+ exit
24
+ fi
25
+
26
+ # start server and pipe stdout+stderr to log file
27
+ cd ~/Work/koboldcpp || exit
28
+ "$PYTHON_EXEC" koboldcpp.py \
29
+ --gpulayers 1 \
30
+ --model "$LLAMA_MODEL_NAME" \
31
+ --contextsize "$LLAMA_CONTEXT_SIZE" \
32
+ --threads 6 \
33
+ --skiplauncher \
34
+ --smartcontext \
35
+ --noblas \
36
+ --host "0.0.0.0" \
37
+ --port "$KOBOLD_PORT" \
38
+ > "$KOBOLD_LOG" 2>&1 &
39
+ echo "Started koboldcpp server on port $KOBOLD_PORT"
bin/llama-update.sh CHANGED
@@ -5,11 +5,12 @@ if [ ! -f "$HOME/.config/llama/llama-update.conf" ]; then
5
  mkdir -p "$HOME/.config/llama"
6
  cat <<EOF > "$HOME/.config/llama/llama-update.conf"
7
  LLAMA_CPP_GIT=$HOME/Work/llama.cpp
 
8
  BIN_PATH=$HOME/.local/bin
9
  EOF
10
  fi
11
 
12
- source $HOME/.config/llama/llama-update.conf
13
 
14
  # if llama.cpp directory does not exist, clone it
15
  if [ ! -d "$LLAMA_CPP_GIT" ]; then
@@ -33,3 +34,23 @@ install -c -v bin/ggml-metal.metal "$BIN_PATH"
33
  install -c -v -m 755 bin/llava-cli "$BIN_PATH"/llava
34
  install -c -v -m 755 bin/finetune "$BIN_PATH/llama-finetune"
35
  install -c -v -m 755 bin/speculative "$BIN_PATH/llama-speculative"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5
  mkdir -p "$HOME/.config/llama"
6
  cat <<EOF > "$HOME/.config/llama/llama-update.conf"
7
  LLAMA_CPP_GIT=$HOME/Work/llama.cpp
8
+ KOBOLDCPP_GIT=$HOME/Work/koboldcpp
9
  BIN_PATH=$HOME/.local/bin
10
  EOF
11
  fi
12
 
13
+ source "$HOME"/.config/llama/llama-update.conf
14
 
15
  # if llama.cpp directory does not exist, clone it
16
  if [ ! -d "$LLAMA_CPP_GIT" ]; then
 
34
  install -c -v -m 755 bin/llava-cli "$BIN_PATH"/llava
35
  install -c -v -m 755 bin/finetune "$BIN_PATH/llama-finetune"
36
  install -c -v -m 755 bin/speculative "$BIN_PATH/llama-speculative"
37
+
38
+ # configured for Metal
39
+ CMAKE_ARGS="-DLLAMA_METAL=on" "$HOME/.virtualenvs/llama.cpp/bin/pip" install \
40
+ 'llama-cpp-python[server]' \
41
+ --force-reinstall \
42
+ --upgrade \
43
+ --no-cache-dir
44
+
45
+ # if koboldcpp directory does not exist, clone it
46
+ if [ ! -d "$KOBOLDCPP_GIT" ]; then
47
+ git clone https://github.com/LostRuins/koboldcpp "$KOBOLDCPP_GIT"
48
+ else
49
+ cd "$KOBOLDCPP_GIT" || exit
50
+ git pull
51
+ fi
52
+
53
+ # update koboldcpp
54
+ cd "$KOBOLDCPP_GIT" || exit
55
+ make clean
56
+ make LLAMA_METAL=1 # LLAMA_CLBLAST=1
bin/llama.sh CHANGED
@@ -8,8 +8,9 @@ function llama_init_environment {
8
  LLAMA_TEMPERATURE=0.1
9
  LLAMA_CONTEXT_SIZE=4096
10
  LLAMA_REPETITION_PENALTY=1.15
11
- LLAMA_TOP_P=0.9
12
- LLAMA_TOP_K=20
 
13
  LLAMA_TEMPLATE=chatml
14
  LLAMA_MODEL_NAME=teknium/OpenHermes-2.5-Mistral-7B/openhermes-2.5-mistral-7b-f16.gguf
15
  LLAMA_MODELS_PATH=$HOME/.ai/models/llama
@@ -52,10 +53,12 @@ function llama_interactive {
52
  --in-suffix "$(get_model_suffix $LLAMA_TEMPLATE)" \
53
  --reverse-prompt "$(get_model_prefix $LLAMA_TEMPLATE)" \
54
  --reverse-prompt "<|im_end|>" \
 
55
  --threads "6" \
56
  --temp "$LLAMA_TEMPERATURE" \
57
  --top-p "$LLAMA_TOP_P" \
58
  --top-k "$LLAMA_TOP_K" \
 
59
  --repeat-penalty "$LLAMA_REPETITION_PENALTY" \
60
  --ctx-size "$LLAMA_CONTEXT_SIZE" \
61
  --batch-size 1024 \
@@ -79,7 +82,7 @@ function get_model_prefix {
79
  *alpaca*|*upstage*|*airoboros*|*hermes*)
80
  printf "### Instruction: "
81
  ;;
82
- *vicuna*|*wizardlm*|*samantha*|*scarlett*)
83
  printf "USER: "
84
  ;;
85
  *based*|*yi*)
@@ -105,7 +108,7 @@ function get_model_suffix {
105
  *alpaca*|*upstage*|*airoboros*|*hermes*)
106
  printf "### Response:"
107
  ;;
108
- *vicuna*|*wizardlm*|*samantha*|*scarlett*)
109
  printf "ASSISTANT:"
110
  ;;
111
  *based*|*yi*)
@@ -149,6 +152,9 @@ function get_model_prompt {
149
  *orca*)
150
  echo "$LLAMA_PROMPT_PATH/orca.txt"
151
  ;;
 
 
 
152
  *)
153
  echo "$LLAMA_PROMPT_PATH/idm-gpt-lite.txt"
154
  ;;
 
8
  LLAMA_TEMPERATURE=0.1
9
  LLAMA_CONTEXT_SIZE=4096
10
  LLAMA_REPETITION_PENALTY=1.15
11
+ LLAMA_TOP_P=1.0
12
+ LLAMA_MIN_P=0.3
13
+ LLAMA_TOP_K=0
14
  LLAMA_TEMPLATE=chatml
15
  LLAMA_MODEL_NAME=teknium/OpenHermes-2.5-Mistral-7B/openhermes-2.5-mistral-7b-f16.gguf
16
  LLAMA_MODELS_PATH=$HOME/.ai/models/llama
 
53
  --in-suffix "$(get_model_suffix $LLAMA_TEMPLATE)" \
54
  --reverse-prompt "$(get_model_prefix $LLAMA_TEMPLATE)" \
55
  --reverse-prompt "<|im_end|>" \
56
+ --reverse-prompt "</s>" \
57
  --threads "6" \
58
  --temp "$LLAMA_TEMPERATURE" \
59
  --top-p "$LLAMA_TOP_P" \
60
  --top-k "$LLAMA_TOP_K" \
61
+ --min-p "$LLAMA_MIN_P" \
62
  --repeat-penalty "$LLAMA_REPETITION_PENALTY" \
63
  --ctx-size "$LLAMA_CONTEXT_SIZE" \
64
  --batch-size 1024 \
 
82
  *alpaca*|*upstage*|*airoboros*|*hermes*)
83
  printf "### Instruction: "
84
  ;;
85
+ *vicuna*|*wizardlm*|*samantha*|*scarlett*|*capybara*)
86
  printf "USER: "
87
  ;;
88
  *based*|*yi*)
 
108
  *alpaca*|*upstage*|*airoboros*|*hermes*)
109
  printf "### Response:"
110
  ;;
111
+ *vicuna*|*wizardlm*|*samantha*|*scarlett*|*capybara*)
112
  printf "ASSISTANT:"
113
  ;;
114
  *based*|*yi*)
 
152
  *orca*)
153
  echo "$LLAMA_PROMPT_PATH/orca.txt"
154
  ;;
155
+ *capybara*)
156
+ echo "$LLAMA_PROMPT_PATH/idm-gpt-lite.txt"
157
+ ;;
158
  *)
159
  echo "$LLAMA_PROMPT_PATH/idm-gpt-lite.txt"
160
  ;;
install.sh CHANGED
@@ -13,6 +13,7 @@ install -C -v ./bin/llama-finetune.sh ~/.local/bin
13
  install -C -v ./bin/llama.sh ~/.local/bin
14
  install -C -v ./bin/llama-menu.sh ~/.local/bin
15
  install -C -v ./bin/llama-server.sh ~/.local/bin
 
16
  install -C -v ./bin/llava-describe.sh ~/.local/bin
17
  install -C -v ./bin/llama-prompt.sh ~/.local/bin
18
 
 
13
  install -C -v ./bin/llama.sh ~/.local/bin
14
  install -C -v ./bin/llama-menu.sh ~/.local/bin
15
  install -C -v ./bin/llama-server.sh ~/.local/bin
16
+ install -C -v ./bin/llama-koboldcpp.sh ~/.local/bin
17
  install -C -v ./bin/llava-describe.sh ~/.local/bin
18
  install -C -v ./bin/llama-prompt.sh ~/.local/bin
19