text stringlengths 0 23.4k |
|---|
Whether you're a hobbyist or a professional, if you want to find the best high-tech drone for your needs, there are a few clear choices. These are some of the highest quality and most advanced drones available for consumer use. |
1. DJI Inspire 1 V2.0 |
The Inspire 1 V2.0 is a high-quality drone built for professionals. It's a massive improvement on DJI's original Inspire 1 drone. It boasts improvements in nearly every area, from the navigation and stability controls to the camera and rotors. However, these improvements do come with an increased price tag. It's built ... |
2. Walkera Voyager 3 |
The Voyager 3 is a drone similar in both its design and functionality to the original DJI Inspire 1. However, it includes features which the Inspire doesn't, such as a visual battery level indicator. As well as additional navigational features. It also costs less than the newer Inspire 1 V2.0. |
3. DJI Phantom 3 |
The Phantom 3 is a versatile mid-range drone available in multiple builds. From the Phantom 3 Standard, which is relatively cheap, and great for hobbyists and beginners to the Phantom 3 Professional, which is over double the price and is packed with more high-tech features. The Phantom 3 quadcopters are great for hobby... |
4. Yuneec Tornado H920 |
The Tornado Hexacopter is a high-quality and robust drone created by Yuneec. While not as high-tech as some other UAVs, the Tornado more than makes up for it with its vastly superior payload capacity and durability. The Tornado has a similar price to the Phantom 1 V2.0. |
5. Walkera QR X900 |
The QR X900 is a six-axis professional quality UAV built for maximum stability. The drone is built with state of the art stability assistance features which can keep the drone stable even during sudden turbulence. It also includes a parachute to minimize damage if the drone loses control. The QR X900 isn't the cheapest... |
6. Yuneec Q500 Typhoon |
The Q500 Typhoon is a great mid-range drone on par with DJI's Phantom series. The Typhoon has many of the capabilities of the Phantom 3, as well as similar pricing. The Typhoon is available in multiple configurations, including a 4K model, and a GoPro compatible model. The Typhoon isn't as accessible for beginners as t... |
These drones and UAVs provide state of the art features and high-tech construction. They range from relatively cheap quadcopters great for enthusiasts and beginners to professional quality hexacopters built to provide extreme stability. These are some of the best high-tech drones you can buy. |
dynamic and static stability assessment of rock slopes ... ?· dynamic and static stability... |
Post on 11-Jun-2018 |
213 views |
0 download |
Embed Size (px) |
Rock Mech. Rock Engng. (2000) 33 (1), 3151Rock Mechanicsand Rock Engineering |
: Springer-Verlag 2000Printed in Austria |
Dynamic and Static Stability Assessment of Rock SlopesAgainst Wedge Failures |
H. Kumsar1, O . Aydan2, and R. Ulusay3 |
1 Pamukkale University, Department of Geological Engineering, Kinikli Campus, Denizli,Turkey |
2 Tokai University, Department of Marine Civil Engineering, Orido, Shimizu, Japan3 Hacettepe University, Department of Geological Engineering, Ankara, Turkey |
The stability of slopes during and after excavation is always of great concern in the eld ofrock engineering. One of the structurally controlled modes of failure in jointed rock slopesis wedge failure. The limiting equilibrium methods for slopes under various conditionsagainst wedge failure have been previously propose... |
1. Introduction |
One of the fundamental problems of engineering geology and geotechnical engi-neering is assessing and maintaining the stability of natural and man-made slopes.Usually natural slopes are stable when they are not disturbed by any externalforce, such as seismic and hydrostatic forces, blasting and surcharge loading.When s... |
Wedge failures are analysed using kinematical approaches and also limiting |
equilibrium methods. The limiting equilibrium method for wedge failure was rstpresented by Wittke (1967) and elaborated for dierent conditions by Kovari andFritz (1975), and Hoek and Bray (1981). Although these methods consider waterpressure, dynamic assessments are not included into the factor of safety computa-tions.... |
2. Preparation of the Wedge Models |
Concrete blocks were used as model materials in this study. Six types of wedgemodel moulds were set up and ordered to be prepared in a plastic factory in orderto obtain smooth surfaces on the wedge models. Since these wedge models weredesigned for testing, and required to be stable at static condition before the tests,... |
Fig. 1. Types of wedge failure |
32 H. Kumsar et al. |
Base and wedge models were formed by using mortar, whose geomechanicalparameters are similar to those of rocks. The composition of the mortar used forthe preparation of the models includes 17.5 kN of ne sand, 3.5 kN of cement and1.75 kN of water. |
In order to model the wedge base and wedge blocks, wooden boxes with alength of 28 cm, a width of 14 cm and a depth of 12 cm were used. The prepara-tion of the concrete models was carried out in two stages consisting of base andwedge model preparations. Base model preparation requires forming two discon-tinuity surface... |
Plastic moulds were extracted from the base models after 8 hours, in order tominimize the damage to them during the extraction, so that discontinuity surfacesfor wedges were obtained. By lling the rest of the space in each box with themortar prepared earlier, the concrete wedge base blocks were modelled. After 24hours,... |
Table 1. Geometric parameters of the prepared model wedges |
Wedge no. Intersectionangle ia0 |
Half wedgeangle oa0 |
Fig. 2. Concrete wedge models used in the study |
Dynamic and Static Stability Assessment 33 |
3. Model Tests |
A model testing program was undertaken to check the validity of the limitingequilibrium method to be presented in the next section. The concrete wedge mod-els were tested under dynamic and static conditions. The static tests were carriedout by considering dry and submerged conditions. The tests were briey describedin t... |
3.1 Static Tests |
The friction angle between the concrete base block and the wedge models wasdetermined from a tilting test, using a portable tilting device. Two concrete blocks,each having the same composition with the base and wedge models, were put oneach other and the tilting angle at which the upper block slides on the lower blocki... |
3.1.1 Dry Tests |
A portable tilting test device (Aydan et al., 1995) was used for the stability tests onthe model wedges. After xing the base blocks on the tilting test device, the wedgeblock models were assembled (Fig. 3). Then the base platform was gradually tiltedby keeping the tilting speed as slow as possible. The inclination angl... |
3.1.2 Submerged Tests |
The stability assessment of the wedge blocks was also carried out under submergedcondition. The wedge base and wedge blocks were put in a water tank, which was |
Fig. 3. Perspective view of the tilting test on a wedge model |
34 H. Kumsar et al. |
lled with water up to a level at which all the model blocks were fully submerged(Fig. 4). The tilting test device was again used to measure the inclination angle ofthe base of the wedge models under submerged condition. Each wedge model wastested three times. For each wedge type, three concrete wedge models were tested... |
3.2 Dynamic Tests |
Rock slope failures may also occur under the inuence of tectonic events in aregion. Heavily jointed rock slopes may indicate that the region was subjected toactive tectonic events in the past. If the tectonic events still continue from time totime, the slopes will be subjected to seismic forces. Even an engineering str... |
Dynamic testing of the wedge models were performed in the laboratory bymeans of a one-dimensional shaking table, which moves along horizontal plane.The wave forms of the shaking table are sinusoidal, saw tooth, rectangular, tra-pezoidal and triangle. The shaking table has a square shape with 1m side length.The frequenc... |
Each wedge base block was xed on the shaking table to receive same shakingwith the shaking table during the dynamic test. The accelerations |
Hello to everyone. |
I have a few hummer plants that I would like to harvest the seeds from, i.e. batfaced, salvias, million bells, snapdragons, etc. but I don't have a clue when is the right time. Does one wait until the flower is completely dried out? Its been raining here and the flowers are not given the chance to dry up. Do I have to ... |
Follow-Up Postings: |
Rudy let the pods dry out in a open container and area. I bring mine inside as the climate is controlled. Don't worry about the pods will split when dried and the seeds will fall out. Then seperate. | |
As a rule million bells don't make seeds but occasionally they will form a seed pod where the flower was after it has dried and fallen off. Seeds develop after the flower has dried and fallen off. salvia seeds can be tricky and it depends on the salvia that you have. Salvia coccineas (lady in red, ect) are easy to coll... |
- Posted by jenny_in_SE_PA USDA7 Sunset 32 (My Page) on Thu, Aug 18, 05 at 11:01 |
Penny - what about B & B seeds? Haven't really checked mine out to see if there are any seeds forming but I guess I should. | |
Jenny, The black and blue seeds have to be watched carefully because the calyx is big and when the seeds are ripe they just fall right out. The calyx will also turn brown when they are ripe. What I do with my potted B&B is to keep a large plant saucer under the pot (largest I can get) and that way any seed that falls i... |
- Posted by jenny_in_SE_PA USDA7 Sunset 32 (My Page) on Thu, Aug 18, 05 at 12:26 |
Thanks for that Penny. My B & B is mostly blooming while suspended in mid air (that's how they grew in order to get to the sun), so if any seeds have formed and fell out, they're way down on the ground or maybe on top of the holly trees that are down below... LOL I do have a few blooming stems that haven't grown throug... |
Jen, That is awsome! You are a real credit to apartment dwellers. I am so glad that you have been successful. Penny | |
- Posted by jenny_in_SE_PA USDA7 Sunset 32 (My Page) on Thu, Aug 18, 05 at 15:57 |
Thanks Penny! :-D Hope to inspire others like so many of you have inspired me. Now to find more room! LOL | |
Jenny There is never enough room. I have dug up more yard space and filled even more containers and planter boxes this year and still don't have enough room. You would think I would quit when it takes most of the day to water the plants. Penny | |
I have been remiss in collecting seeds -- and I had such good intentions this spring! | |
- Posted by RubyHummerLover z5CO (My Page) on Fri, Aug 19, 05 at 8:38 |
Thanks to all for the great tips. I will be starting my harvesting this weekend. Penny 1947, thanks for informing me about the million bells, I thought they were just tiny petunias. They are so cute. I will definitely be buying more of those next season. Here's to everyone having a great hummer weekend!!! Happy harvest... |
Oubliette, You should still have plenty of time to start collecting. I don't start until mid to late July once I know that the flowers are being visited and pollinated. Penny | |
Thanks, Penny. Is it too late to collect the se The only other things that have bloom Salvia B & B doesn't come back true from se What about pentas? How do you collect those se Thanks!! | |
If the Penny | |
Thanks Penny -- will try this out. Disclaimer: I do not agree with nor endorse the use of "Sponsored Links" possibly embedded in my posts by the owner of this or any other website and would never purchase anything from any advertiser that used this deceptive type of advertising. | |
Please Note: Only registered members are able to post messages to this forum. If you are a member, please log in. If you aren't yet a member, join now! |
Return to the Hummingbird Garden Forum |
Information about Posting |
- You must be logged in to post a message. Once you are logged in, a posting window will appear at the bottom of the messages. If you are not a member, please register for an account. |
- Please review our Rules of Play before posting. |
- Posting is a two-step process. Once you have composed your message, you will be taken to the preview page. You will then have a chance to review your post, make changes and upload photos. |
- After posting your message, you may need to refresh the forum page in order to see it. |
- Before posting copyrighted material, please read about Copyright and Fair Use. |
- We have a strict no-advertising policy! |
- If you would like to practice posting or uploading photos, please visit our Test forum. |
- If you need assistance, please Contact Us and we will be happy to help. |
Learn more about in-text links on this page here |
Doctor insights on: |
Beta Carotene Side Effects |
Beta blocker side effects? |
Poss side effects: Possible side effects of beta-blockers include: sexual dysfunction in males, HDL reduction, increased triglycerides, bronchoconstriction/ shortness of breath, bradycardia, hypotension, heart block, fatigue, and dizziness. However, don't forget that most people do fine, or have a brief adjustment peri... |
Beta blocker side effect? |
Fatigue: Fatigue is the most common side effect, which may include daytime sleepiness. Erectile dysfunction may also occur. A beta- blocker also lowers heart rate and blood pressure, so if you are sensitive to it, or the dose is too large, bradycardia (low heart rate) or hypotension (low blood pressure) may occur. ...R... |
Any side effects to beta carotene v tocopherol? |
K2-Horizon-MoVA-36B-A4B: BF16 reference log-probs
Full-vocabulary log-softmax outputs of IFM/K2-Horizon-MoVA-36B-A4B in BF16 (run in MLX). They are the reference for
KL(P_bf16 || P_candidate) on two splits. With these files, a quantized or compressed engine can be scored on any
platform without the 75 GB BF16 checkpoint or MLX.
These are the files behind the NSLMv2 Mac KLD tables, byte for byte. They are not a regeneration.
Layout
Each split has CTX = 2048 windows: BOS followed by 2047 text tokens, with no overlap. Rows 1024..2046 of each window
predict tokens 1025..2047, so each window has 1023 scored positions. Arrays are little-endian.
| file | dtype / shape | contents |
|---|---|---|
{split}/ref.f16.npy |
f16 [positions][250624] |
log-softmax over the full vocab (NumPy v1 header, then raw data) |
{split}/ref.meta.npz |
npz | reference NLL and argmax per position (Python side) |
{split}/windows.ids |
int32 [windows][2048] |
input token IDs. Use these, not a re-tokenisation |
{split}/windows.refnll |
f32 [positions] |
reference NLL of the true next token |
{split}/windows.reftop |
int32 [positions] |
reference argmax |
heldout/windows.routes_mlp |
int32 [w][2047][45][8] |
MLX BF16 MLP-expert top-8 per sparse layer, sorted |
heldout/windows.routes_val |
int32 [w][2047][45][4] |
MLX BF16 value-expert top-4, sorted |
heldout/windows.gap_mlp, gap_val |
f32 [w][2047][45] |
k-th minus (k+1)-th router score (near-tie test, < 1e-3) |
text/* |
text | source texts the windows were tokenised from |
SHA256SUMS |
shasum -a 256 -c SHA256SUMS |
| split | windows | positions | ref.f16.npy | ref PPL |
|---|---|---|---|---|
heldout |
31 | 31,713 | 15.9 GB | 10.7579 |
chat |
29 | 29,667 | 14.9 GB | 1.5463 |
Metric
For each position: renormalise the stored log-probs (r = lp - logsumexp(lp)), then take the exact full-vocab
KL = sum_v exp(r_v) * (r_v - log q_v) against the candidate's log-softmax q. Report the mean over positions, its
standard error, the PPL ratio and top-1 agreement. The reference implementation is harness/nslm-mova-score.m. Its
inputs map as --windows {split}/windows and --ref {split}/ref. The candidate engine still needs the model's
config.json from IFM/K2-Horizon-MoVA-36B-A4B.
Mac anchors (Apple M4 Max) for checking a port:
| candidate | heldout KLD | heldout top-1 | chat KLD | chat top-1 |
|---|---|---|---|---|
our engine, all BF16 (bf16-c) |
0.0054920 | 96.705% | ||
| MLX BF16 vs reference (floor) | 0.0051923 | |||
| q8mx | 0.0072955 | 96.128% | 0.0050159 | 98.733% |
| nslmmx | 0.0345328 | 91.921% | 0.0137807 | 97.172% |
Provenance
- Model:
IFM/K2-Horizon-MoVA-36B-A4B@cca48b6631d03c338a0e1d8beb0c392af8becbd3(Apache-2.0), BF16 tokenizer.jsonsha2562fa69519ff1ee3ee55ac50927b00058e88b9eb90802fabaf3c6b05804601c53bchat_template.jinjasha256a892cd0b0195599f283a8c706787520d9a6747640efb2f4dec4144b0abb62590- Runtime: MLX 0.32.2, mlx-lm 0.31.3, Python 3.13, macOS, written 2026-10-04
- Generator: NanoSeedLM
tools/mova_score.py --write-refandtools/mova_export.py windows|routes(research-repo commitsc1a454d,5cd1284) heldouttext: 68 documents fromIFM/TxT360-v2@a86bdfb1,web-high-medium/Medium-High.chunk0-9fc7d32df3-00000.parquet, row group 80. The selection is intext/heldout_manifest.json. TxT360-v2 is licensed CC-BY-4.0, and this redistribution keeps that license.chattext: MoVA BF16's own greedy and sampled responses to the NSLMv2 generation-suite prompts, formatted with the model's chat template (tools/make_chat_text.py)
License
CC-BY-4.0. The held-out text and the log-probs derived from it are attributed to IFM's TxT360-v2. The model outputs come from an Apache-2.0 model.
- Downloads last month
- 93