Skip to content

Commit 5b9b86e

Browse files
authored
Update README.md
1 parent 9c1f716 commit 5b9b86e

File tree

1 file changed

+15
-15
lines changed

1 file changed

+15
-15
lines changed

README.md

Lines changed: 15 additions & 15 deletions
Original file line numberDiff line numberDiff line change
@@ -25,7 +25,7 @@ To restore a checkpoint, you must specify the path to a folder that has "actor.p
2525

2626
* Uses a GRU instead of LSTM for the decoder network
2727
* Critic takes the raw static and dynamic input states and predicts a reward
28-
* Here we use demand scaling (MAX_DEMAND / MAX_VEHICLE_CAPACITY), and give the depot for the VRP a negative value proportionaly to the missing capacity (Unsure if used or not)
28+
* Use demand scaling (MAX_DEMAND / MAX_VEHICLE_CAPACITY), and give the depot for the VRP a negative value proportionate to the missing capacity (Unsure if used or not)
2929

3030
## TSP Sample Tours:
3131

@@ -89,7 +89,7 @@ This repo only implements the "Greedy" approach during test time, which selects
8989

9090
## Training Time
9191

92-
On a Tesla P-100 GPU, the following training times are observed. Results were obtained by taking the the total time for the first 100 training iterations (with respective batch sizes), and converting into the appopriate time unit. Note that for the VRP in particular, as models are relatively untrained during this time, this may be slightly inaccurating results and YMMV.
92+
On a Tesla P-100 GPU, the following training times are observed. Results were obtained by taking the the total time for the first 100 training iterations (with respective batch sizes), and converting into the appopriate time unit. Note that for the VRP in particular, as models are relatively untrained during this time, this may be slightly inaccurate results and YMMV.
9393

9494
| Task | Batch Size | Sec / 100 Updates | Min / Epoch | Hours/Epoch | 20 Epochs |
9595
|--------|------------|-------------------|-------------|-------------|-----------|
@@ -102,19 +102,19 @@ On a Tesla P-100 GPU, the following training times are observed. Results were ob
102102
| TSP100 | 128 | 48.27 | 62.85 | 1.05 | 20.95 |
103103
| TSP100 | 256 | 73.51 | 47.85 | 0.80 | 15.95 |
104104

105-
| Task | Batch Size | Time / 100 Updates | Min / Epoch | Hours/Epoch | 20 Epochs |
106-
|--------|------------|--------------------|-------------|-------------|-----------|
107-
| VRP10 | 128 | 12.15 | 15.82 | 0.26 | 5.27 |
108-
| VRP10 | 256 | 15.75 | 10.25 | 0.17 | 3.42 |
109-
| VRP10 | 512 | 23.30 | 7.58 | 0.13 | 2.53 |
110-
| VRP20 | 128 | 21.45 | 27.93 | 0.47 | 9.31 |
111-
| VRP20 | 256 | 28.29 | 18.42 | 0.31 | 6.14 |
112-
| VRP20 | 512 | 43.20 | 14.06 | 0.23 | 4.69 |
113-
| VRP50 | 128 | 53.59 | 69.77 | 1.16 | 23.26 |
114-
| VRP50 | 256 | 77.25 | 50.29 | 0.84 | 16.76 |
115-
| VRP50 | 512 | 127.73 | 41.58 | 0.69 | 13.86 |
116-
| VRP100 | 128 | 130.06 | 169.35 | 2.82 | 56.45 |
117-
| VRP100 | 64 | 95.03 | 247.48 | 4.12 | 82.49 |
105+
| Task | Batch Size | Sec / 100 Updates | Min / Epoch | Hours/Epoch | 20 Epochs |
106+
|--------|------------|-------------------|-------------|-------------|-----------|
107+
| VRP10 | 128 | 12.15 | 15.82 | 0.26 | 5.27 |
108+
| VRP10 | 256 | 15.75 | 10.25 | 0.17 | 3.42 |
109+
| VRP10 | 512 | 23.30 | 7.58 | 0.13 | 2.53 |
110+
| VRP20 | 128 | 21.45 | 27.93 | 0.47 | 9.31 |
111+
| VRP20 | 256 | 28.29 | 18.42 | 0.31 | 6.14 |
112+
| VRP20 | 512 | 43.20 | 14.06 | 0.23 | 4.69 |
113+
| VRP50 | 128 | 53.59 | 69.77 | 1.16 | 23.26 |
114+
| VRP50 | 256 | 77.25 | 50.29 | 0.84 | 16.76 |
115+
| VRP50 | 512 | 127.73 | 41.58 | 0.69 | 13.86 |
116+
| VRP100 | 128 | 130.06 | 169.35 | 2.82 | 56.45 |
117+
| VRP100 | 64 | 95.03 | 247.48 | 4.12 | 82.49 |
118118

119119
# Acknowledgements:
120120

0 commit comments

Comments
 (0)