Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions docker/npu_patch/slime.patch
Original file line number Diff line number Diff line change
Expand Up @@ -692,7 +692,7 @@ index 9f51ecdf..d4b47eca 100644
+ if not external_ray:
+ exec_command(
+ # will prevent ray from buffering stdout/stderr
+ f"export PYTHONBUFFERED=16 && "
+ f"export PYTHONUNBUFFERED=1 && "
+ f"ray start --head --node-ip-address {master_addr} --disable-usage-stats"
+ )
+
Expand Down Expand Up @@ -748,7 +748,7 @@ index 9f51ecdf..d4b47eca 100644
+ else ""
+ )
+ exec_command(
+ f"export no_proxy=127.0.0.1 && export PYTHONBUFFERED=16 && "
+ f"export no_proxy=127.0.0.1 && export PYTHONUNBUFFERED=1 && "
+ f"{cmd_megatron_model_source}"
+ f'ray job submit --address="http://127.0.0.1:8265" '
+ f"--runtime-env-json='{runtime_env_json}' "
Expand Down
2 changes: 1 addition & 1 deletion docs/en/platform_support/amd_tutorial.md
Original file line number Diff line number Diff line change
Expand Up @@ -148,7 +148,7 @@ export HIP_VISIBLE_DEVICES=${HIP_VISIBLE_DEVICES:-"0,1,2,3,4,5,6,7"} #You can ch


# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
source "${SCRIPT_DIR}/models/qwen3-4B.sh"
Expand Down
2 changes: 1 addition & 1 deletion examples/delta_weight_sync/run-glm4.7-355B-A32B-delta.sh
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@ pkill -9 python

set -ex

export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1
unset http_proxy https_proxy HTTP_PROXY HTTPS_PROXY

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
Expand Down
2 changes: 1 addition & 1 deletion examples/eval_multi_task/multi_task.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion examples/fully_async/run-qwen2.5-0.5B-fully_async.sh
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ sleep 3

set -ex

export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
HAS_NVLINK=$([ "$NVLINK_COUNT" -gt 0 ] && echo 1 || echo 0)
Expand Down
2 changes: 1 addition & 1 deletion examples/geo3k_vlm/run_geo3k_qwen35.sh
Original file line number Diff line number Diff line change
Expand Up @@ -44,7 +44,7 @@ pkill -9 redis

set -ex

export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

# Detect NVLink
NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
Expand Down
2 changes: 1 addition & 1 deletion examples/geo3k_vlm/run_geo3k_vlm.sh
Original file line number Diff line number Diff line change
Expand Up @@ -59,7 +59,7 @@ pkill -9 redis

set -ex

export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

# Detect NVLink
NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
Expand Down
2 changes: 1 addition & 1 deletion examples/geo3k_vlm/run_geo3k_vlm_sft.sh
Original file line number Diff line number Diff line change
Expand Up @@ -52,7 +52,7 @@ pkill -9 redis

set -ex

export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

# Detect NVLink
NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
Expand Down
2 changes: 1 addition & 1 deletion examples/multi_agent/run-qwen3-30B-A3B-multi-agent.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@

set -ex

export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion examples/on_policy_distillation/run-qwen3-8B-opd.sh
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@ echo "Teacher model server is up and running at $TEACHER_IP:$TEACHER_PORT."
sleep 10


export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion examples/retool/retool_qwen3_4b_rl.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion examples/retool/retool_qwen3_4b_sft.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion examples/search-r1/run_qwen2.5_3B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
source "${SCRIPT_DIR}/../../scripts/models/qwen2.5-3B.sh"
Expand Down
2 changes: 1 addition & 1 deletion examples/strands_sglang/strands_qwen3_8b.sh
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion examples/tau-bench/run_qwen3_4B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion examples/train_infer_mismatch_helper/run-qwen3-4b-mis.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/low_precision/run-kimi-k2-Thinking-int4.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi | grep -o "NVLink" | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/low_precision/run-moonlight-16B-A3B-int4.sh
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,7 @@ pkill -9 redis
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/low_precision/run-qwen3-235B-A22B-int4.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi | grep -o "NVLink" | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
4 changes: 2 additions & 2 deletions scripts/low_precision/run-qwen3-30B-A3B-int4.sh
Original file line number Diff line number Diff line change
Expand Up @@ -12,8 +12,8 @@ pkill -9 python

set -ex

# will prevent ray from buffering stdout/stderrs
export PYTHONBUFFERED=16
# will prevent ray from buffering stdout/stderr
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/low_precision/run-qwen3-30b-a3b-fp8.sh
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,7 @@
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/low_precision/run-qwen3-4b-fp8.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-deepseek-r1.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-glm4-9B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-glm4.7-30B-A3B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

unset http_proxy https_proxy HTTP_PROXY HTTPS_PROXY

Expand Down
2 changes: 1 addition & 1 deletion scripts/run-glm4.7-355B-A32B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

unset http_proxy https_proxy HTTP_PROXY HTTPS_PROXY

Expand Down
2 changes: 1 addition & 1 deletion scripts/run-glm5-744B-A40B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-gpt-oss-20B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@ pkill -9 python

set -ex

export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-kimi-k2-Instruct.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-kimi-k2-Thinking.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-mimo-7B-rl-eagle.sh
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-moonlight-16B-A3B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,7 @@ pkill -9 redis
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen2.5-0.5B-gb10-smoke.sh
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ ray stop --force 2>/dev/null || true
pkill -9 ray python 2>/dev/null || true
sleep 2

export PYTHONBUFFERED=1
export PYTHONUNBUFFERED=1

SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
source "${SCRIPT_DIR}/models/qwen2.5-0.5B.sh"
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen2.5-0.5B-reproducibility.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
source "${SCRIPT_DIR}/scripts/models/qwen2.5-0.5B.sh"
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-235B-A22B-sft.sh
Original file line number Diff line number Diff line change
Expand Up @@ -24,7 +24,7 @@ if [ -z "${MASTER_ADDR}" ]; then
fi

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-235B-A22B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -24,7 +24,7 @@ if [ -z "${MASTER_ADDR}" ]; then
fi

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-30B-A3B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,7 @@ pkill -9 redis
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-32B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-4B-amd.sh
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ export HIP_VISIBLE_DEVICES=${HIP_VISIBLE_DEVICES:-"0,1,2,3,4,5,6,7"} #You can ch


# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
source "${SCRIPT_DIR}/models/qwen3-4B.sh"
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-4B-base-sft.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-4B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ pkill -9 python
set -ex

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

NVLINK_COUNT=$(nvidia-smi topo -m 2>/dev/null | grep -o 'NV[0-9][0-9]*' | wc -l)
if [ "$NVLINK_COUNT" -gt 0 ]; then
Expand Down
2 changes: 1 addition & 1 deletion scripts/run-qwen3-next-80B-A3B.sh
Original file line number Diff line number Diff line change
Expand Up @@ -24,7 +24,7 @@ if [ -z "${MASTER_ADDR}" ]; then
fi

# will prevent ray from buffering stdout/stderr
export PYTHONBUFFERED=16
export PYTHONUNBUFFERED=1

# unset proxy to avoid distributed startup issues
unset http_proxy https_proxy HTTP_PROXY HTTPS_PROXY
Expand Down
Loading